<?xml version="1.0" encoding="utf-8"?>
<feed xmlns="http://www.w3.org/2005/Atom"><id>tag:odin-labs-ai.github.io,2026:odin-rnd/feed</id><title>Odin Labs R&amp;D: every update and every lab run</title><subtitle>Experiment plans, results, corrections, tools and every run of every lab&apos;s check, newest first, from the records this site publishes.</subtitle><updated>2026-11-17T05:39:13Z</updated><link rel="self" type="application/atom+xml" href="https://odin-labs-ai.github.io/odin-rnd/feed.xml"/><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/updates/"/><author><name>Odin Labs R&amp;D</name></author><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-17T05:39:13Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on ubuntu-latest, record 04fa0d5ba8de, SAME</title><updated>2026-11-17T05:39:13Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261117053913-04fa0d5ba8de"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-17T05:39:05Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on org-ci-heavy, record 04fa0d5ba8de, SAME</title><updated>2026-11-17T05:39:05Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261117053905-04fa0d5ba8de"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-17T05:25:51Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on ubuntu-latest, record f2c48fc2a28a, SAME</title><updated>2026-11-17T05:25:51Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261117052551-f2c48fc2a28a"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-17T05:25:39Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on org-ci-heavy, record f2c48fc2a28a, SAME</title><updated>2026-11-17T05:25:39Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261117052539-f2c48fc2a28a"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-16T05:39:45Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on ubuntu-latest, record 04fa0d5ba8de, SAME</title><updated>2026-11-16T05:39:45Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261116053945-04fa0d5ba8de"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-16T05:39:43Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on org-ci-heavy, record 04fa0d5ba8de, SAME</title><updated>2026-11-16T05:39:43Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261116053943-04fa0d5ba8de"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-16T05:28:28Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on ubuntu-latest, record f2c48fc2a28a, SAME</title><updated>2026-11-16T05:28:28Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261116052828-f2c48fc2a28a"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-16T05:28:15Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on org-ci-heavy, record f2c48fc2a28a, SAME</title><updated>2026-11-16T05:28:15Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261116052815-f2c48fc2a28a"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-15T05:39:09Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on org-ci-heavy, record 04fa0d5ba8de, SAME</title><updated>2026-11-15T05:39:09Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261115053909-04fa0d5ba8de"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-15T05:39:01Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on ubuntu-latest, record 04fa0d5ba8de, SAME</title><updated>2026-11-15T05:39:01Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261115053901-04fa0d5ba8de"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-15T05:26:00Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on ubuntu-latest, record f2c48fc2a28a, SAME</title><updated>2026-11-15T05:26:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261115052600-f2c48fc2a28a"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-15T05:25:46Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on org-ci-heavy, record f2c48fc2a28a, SAME</title><updated>2026-11-15T05:25:46Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261115052546-f2c48fc2a28a"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-14T05:39:33Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on org-ci-heavy, record 04fa0d5ba8de, SAME</title><updated>2026-11-14T05:39:33Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261114053933-04fa0d5ba8de"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-14T05:38:39Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on ubuntu-latest, record 04fa0d5ba8de, SAME</title><updated>2026-11-14T05:38:39Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261114053839-04fa0d5ba8de"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-14T05:26:11Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on org-ci-heavy, record f2c48fc2a28a, SAME</title><updated>2026-11-14T05:26:11Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261114052611-f2c48fc2a28a"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-14T05:25:31Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on ubuntu-latest, record f2c48fc2a28a, SAME</title><updated>2026-11-14T05:25:31Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261114052531-f2c48fc2a28a"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-13T19:32:47Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on org-ci-heavy, record f2c48fc2a28a, SAME</title><updated>2026-11-13T19:32:47Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261113193247-f2c48fc2a28a"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-13T19:04:48Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on ubuntu-latest, record f2c48fc2a28a, SAME</title><updated>2026-11-13T19:04:48Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261113190448-f2c48fc2a28a"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/org-ci-heavy/2026-11-13T07:44:29Z/f419d9215e2b4f5cc1a5c20f3cd7ba90af773d6844c37537a5424fc15aed33fa</id><title>Lab 13: a run on org-ci-heavy, record f419d9215e2b, SAME</title><updated>2026-11-13T07:44:29Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-org-ci-heavy-20261113074429-f419d9215e2b"/><summary>Does an agent company get better, measured blind? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash f419d9215e2b4f5cc1a5c20f3cd7ba90af773d6844c37537a5424fc15aed33fa; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/ubuntu-latest/2026-11-13T07:39:48Z/f419d9215e2b4f5cc1a5c20f3cd7ba90af773d6844c37537a5424fc15aed33fa</id><title>Lab 13: a run on ubuntu-latest, record f419d9215e2b, SAME</title><updated>2026-11-13T07:39:48Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-ubuntu-latest-20261113073948-f419d9215e2b"/><summary>Does an agent company get better, measured blind? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash f419d9215e2b4f5cc1a5c20f3cd7ba90af773d6844c37537a5424fc15aed33fa; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/org-ci-heavy/2026-11-13T07:14:39Z/9e1467d6b2c8ee72ba76d70afd90b2e4110dd695f1671c92d1133ee9f182d7c5</id><title>Lab 08: a run on org-ci-heavy, record 9e1467d6b2c8, SAME</title><updated>2026-11-13T07:14:39Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-org-ci-heavy-20261113071439-9e1467d6b2c8"/><summary>Can an agent fix a red build? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash 9e1467d6b2c8ee72ba76d70afd90b2e4110dd695f1671c92d1133ee9f182d7c5; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/ubuntu-latest/2026-11-13T07:05:56Z/9e1467d6b2c8ee72ba76d70afd90b2e4110dd695f1671c92d1133ee9f182d7c5</id><title>Lab 08: a run on ubuntu-latest, record 9e1467d6b2c8, SAME</title><updated>2026-11-13T07:05:56Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-ubuntu-latest-20261113070556-9e1467d6b2c8"/><summary>Can an agent fix a red build? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash 9e1467d6b2c8ee72ba76d70afd90b2e4110dd695f1671c92d1133ee9f182d7c5; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-13T05:39:04Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on ubuntu-latest, record 04fa0d5ba8de, SAME</title><updated>2026-11-13T05:39:04Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261113053904-04fa0d5ba8de"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-13T05:39:00Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on org-ci-heavy, record 04fa0d5ba8de, SAME</title><updated>2026-11-13T05:39:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261113053900-04fa0d5ba8de"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/ubuntu-latest/2026-11-12T05:55:50Z/f419d9215e2b4f5cc1a5c20f3cd7ba90af773d6844c37537a5424fc15aed33fa</id><title>Lab 13: a run on ubuntu-latest, record f419d9215e2b, SAME</title><updated>2026-11-12T05:55:50Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-ubuntu-latest-20261112055550-f419d9215e2b"/><summary>Does an agent company get better, measured blind? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash f419d9215e2b4f5cc1a5c20f3cd7ba90af773d6844c37537a5424fc15aed33fa; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/org-ci-heavy/2026-11-12T05:55:17Z/f419d9215e2b4f5cc1a5c20f3cd7ba90af773d6844c37537a5424fc15aed33fa</id><title>Lab 13: a run on org-ci-heavy, record f419d9215e2b, SAME</title><updated>2026-11-12T05:55:17Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-org-ci-heavy-20261112055517-f419d9215e2b"/><summary>Does an agent company get better, measured blind? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash f419d9215e2b4f5cc1a5c20f3cd7ba90af773d6844c37537a5424fc15aed33fa; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/ubuntu-latest/2026-11-12T05:49:18Z/9e1467d6b2c8ee72ba76d70afd90b2e4110dd695f1671c92d1133ee9f182d7c5</id><title>Lab 08: a run on ubuntu-latest, record 9e1467d6b2c8, SAME</title><updated>2026-11-12T05:49:18Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-ubuntu-latest-20261112054918-9e1467d6b2c8"/><summary>Can an agent fix a red build? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash 9e1467d6b2c8ee72ba76d70afd90b2e4110dd695f1671c92d1133ee9f182d7c5; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/org-ci-heavy/2026-11-12T05:48:43Z/9e1467d6b2c8ee72ba76d70afd90b2e4110dd695f1671c92d1133ee9f182d7c5</id><title>Lab 08: a run on org-ci-heavy, record 9e1467d6b2c8, SAME</title><updated>2026-11-12T05:48:43Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-org-ci-heavy-20261112054843-9e1467d6b2c8"/><summary>Can an agent fix a red build? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash 9e1467d6b2c8ee72ba76d70afd90b2e4110dd695f1671c92d1133ee9f182d7c5; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-12T05:38:53Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on ubuntu-latest, record 04fa0d5ba8de, SAME</title><updated>2026-11-12T05:38:53Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261112053853-04fa0d5ba8de"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-12T05:38:37Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on org-ci-heavy, record 04fa0d5ba8de, SAME</title><updated>2026-11-12T05:38:37Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261112053837-04fa0d5ba8de"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-12T05:25:47Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on ubuntu-latest, record f2c48fc2a28a, SAME</title><updated>2026-11-12T05:25:47Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261112052547-f2c48fc2a28a"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-12T05:25:33Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on org-ci-heavy, record f2c48fc2a28a, SAME</title><updated>2026-11-12T05:25:33Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261112052533-f2c48fc2a28a"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/ubuntu-latest/2026-11-11T05:56:34Z/f419d9215e2b4f5cc1a5c20f3cd7ba90af773d6844c37537a5424fc15aed33fa</id><title>Lab 13: a run on ubuntu-latest, record f419d9215e2b, SAME</title><updated>2026-11-11T05:56:34Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-ubuntu-latest-20261111055634-f419d9215e2b"/><summary>Does an agent company get better, measured blind? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash f419d9215e2b4f5cc1a5c20f3cd7ba90af773d6844c37537a5424fc15aed33fa; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/org-ci-heavy/2026-11-11T05:56:07Z/f419d9215e2b4f5cc1a5c20f3cd7ba90af773d6844c37537a5424fc15aed33fa</id><title>Lab 13: a run on org-ci-heavy, record f419d9215e2b, SAME</title><updated>2026-11-11T05:56:07Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-org-ci-heavy-20261111055607-f419d9215e2b"/><summary>Does an agent company get better, measured blind? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash f419d9215e2b4f5cc1a5c20f3cd7ba90af773d6844c37537a5424fc15aed33fa; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/ubuntu-latest/2026-11-11T05:49:02Z/9e1467d6b2c8ee72ba76d70afd90b2e4110dd695f1671c92d1133ee9f182d7c5</id><title>Lab 08: a run on ubuntu-latest, record 9e1467d6b2c8, SAME</title><updated>2026-11-11T05:49:02Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-ubuntu-latest-20261111054902-9e1467d6b2c8"/><summary>Can an agent fix a red build? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash 9e1467d6b2c8ee72ba76d70afd90b2e4110dd695f1671c92d1133ee9f182d7c5; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/org-ci-heavy/2026-11-11T05:48:28Z/9e1467d6b2c8ee72ba76d70afd90b2e4110dd695f1671c92d1133ee9f182d7c5</id><title>Lab 08: a run on org-ci-heavy, record 9e1467d6b2c8, SAME</title><updated>2026-11-11T05:48:28Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-org-ci-heavy-20261111054828-9e1467d6b2c8"/><summary>Can an agent fix a red build? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash 9e1467d6b2c8ee72ba76d70afd90b2e4110dd695f1671c92d1133ee9f182d7c5; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-11T05:39:32Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on ubuntu-latest, record 04fa0d5ba8de, SAME</title><updated>2026-11-11T05:39:32Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261111053932-04fa0d5ba8de"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-11T05:39:27Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on org-ci-heavy, record 04fa0d5ba8de, SAME</title><updated>2026-11-11T05:39:27Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261111053927-04fa0d5ba8de"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-11T05:25:59Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on ubuntu-latest, record f2c48fc2a28a, SAME</title><updated>2026-11-11T05:25:59Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261111052559-f2c48fc2a28a"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-11T05:25:40Z/f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101</id><title>Lab 05: a run on org-ci-heavy, record f2c48fc2a28a, SAME</title><updated>2026-11-11T05:25:40Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261111052540-f2c48fc2a28a"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit c169ad9b; record hash f2c48fc2a28a5a6765be0bd45623e6a556b84f8e66e562509eff3a2ddb449101; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-021-amendment-03</id><title>Amendment · EXP 021: Kill the database mid-session, does anyone notice?</title><updated>2026-11-11T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/failover-results.html"/><summary>The counted run is complete: 48 discards reconciled; strict acknowledgements and one disturbance window remain unresolved.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-021-results</id><title>Results · EXP 021: Kill the database mid-session, does anyone notice?</title><updated>2026-11-11T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/failover-results.html"/><summary>Single host: MongoDB fits four write-outage and two acknowledgement-gap scenarios; PostgreSQL two by feed gate. No acknowledged writes lost; no overall winner. Strict-ack and disturbance uncertainties remain.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/ubuntu-latest/2026-11-10T05:55:55Z/f6b457a9cb814d7aaf8734dae552e6640890e0517d78a1b516788f278db0d67f</id><title>Lab 13: a run on ubuntu-latest, record f6b457a9cb81, SAME</title><updated>2026-11-10T05:55:55Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-ubuntu-latest-20261110055555-f6b457a9cb81"/><summary>Does an agent company get better, measured blind? The github-hosted runner, machine class ubuntu-latest, ran commit d934f799; record hash f6b457a9cb814d7aaf8734dae552e6640890e0517d78a1b516788f278db0d67f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/org-ci-heavy/2026-11-10T05:55:23Z/f6b457a9cb814d7aaf8734dae552e6640890e0517d78a1b516788f278db0d67f</id><title>Lab 13: a run on org-ci-heavy, record f6b457a9cb81, SAME</title><updated>2026-11-10T05:55:23Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-org-ci-heavy-20261110055523-f6b457a9cb81"/><summary>Does an agent company get better, measured blind? The org-hosted runner, machine class org-ci-heavy, ran commit d934f799; record hash f6b457a9cb814d7aaf8734dae552e6640890e0517d78a1b516788f278db0d67f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/ubuntu-latest/2026-11-10T05:49:06Z/a964b2f3fc74fba04a437b9ed04fa1bfae1b114f3a0dbff839fc874bf73ba958</id><title>Lab 08: a run on ubuntu-latest, record a964b2f3fc74, SAME</title><updated>2026-11-10T05:49:06Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-ubuntu-latest-20261110054906-a964b2f3fc74"/><summary>Can an agent fix a red build? The github-hosted runner, machine class ubuntu-latest, ran commit d934f799; record hash a964b2f3fc74fba04a437b9ed04fa1bfae1b114f3a0dbff839fc874bf73ba958; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/org-ci-heavy/2026-11-10T05:48:41Z/a964b2f3fc74fba04a437b9ed04fa1bfae1b114f3a0dbff839fc874bf73ba958</id><title>Lab 08: a run on org-ci-heavy, record a964b2f3fc74, SAME</title><updated>2026-11-10T05:48:41Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-org-ci-heavy-20261110054841-a964b2f3fc74"/><summary>Can an agent fix a red build? The org-hosted runner, machine class org-ci-heavy, ran commit d934f799; record hash a964b2f3fc74fba04a437b9ed04fa1bfae1b114f3a0dbff839fc874bf73ba958; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-10T05:38:27Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on ubuntu-latest, record 04fa0d5ba8de, SAME</title><updated>2026-11-10T05:38:27Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261110053827-04fa0d5ba8de"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-10T05:38:09Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on org-ci-heavy, record 04fa0d5ba8de, SAME</title><updated>2026-11-10T05:38:09Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261110053809-04fa0d5ba8de"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-10T05:25:49Z/feb80a48b4596e6235b737288bcb0c87080354cb42411307a7acf9bc22de2827</id><title>Lab 05: a run on ubuntu-latest, record feb80a48b459, SAME</title><updated>2026-11-10T05:25:49Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261110052549-feb80a48b459"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit d934f799; record hash feb80a48b4596e6235b737288bcb0c87080354cb42411307a7acf9bc22de2827; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-10T05:25:35Z/feb80a48b4596e6235b737288bcb0c87080354cb42411307a7acf9bc22de2827</id><title>Lab 05: a run on org-ci-heavy, record feb80a48b459, SAME</title><updated>2026-11-10T05:25:35Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261110052535-feb80a48b459"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit d934f799; record hash feb80a48b4596e6235b737288bcb0c87080354cb42411307a7acf9bc22de2827; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/ubuntu-latest/2026-11-09T05:56:00Z/f6b457a9cb814d7aaf8734dae552e6640890e0517d78a1b516788f278db0d67f</id><title>Lab 13: a run on ubuntu-latest, record f6b457a9cb81, SAME</title><updated>2026-11-09T05:56:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-ubuntu-latest-20261109055600-f6b457a9cb81"/><summary>Does an agent company get better, measured blind? The github-hosted runner, machine class ubuntu-latest, ran commit d934f799; record hash f6b457a9cb814d7aaf8734dae552e6640890e0517d78a1b516788f278db0d67f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/org-ci-heavy/2026-11-09T05:55:34Z/f6b457a9cb814d7aaf8734dae552e6640890e0517d78a1b516788f278db0d67f</id><title>Lab 13: a run on org-ci-heavy, record f6b457a9cb81, SAME</title><updated>2026-11-09T05:55:34Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-org-ci-heavy-20261109055534-f6b457a9cb81"/><summary>Does an agent company get better, measured blind? The org-hosted runner, machine class org-ci-heavy, ran commit d934f799; record hash f6b457a9cb814d7aaf8734dae552e6640890e0517d78a1b516788f278db0d67f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/ubuntu-latest/2026-11-09T05:49:00Z/a964b2f3fc74fba04a437b9ed04fa1bfae1b114f3a0dbff839fc874bf73ba958</id><title>Lab 08: a run on ubuntu-latest, record a964b2f3fc74, SAME</title><updated>2026-11-09T05:49:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-ubuntu-latest-20261109054900-a964b2f3fc74"/><summary>Can an agent fix a red build? The github-hosted runner, machine class ubuntu-latest, ran commit d934f799; record hash a964b2f3fc74fba04a437b9ed04fa1bfae1b114f3a0dbff839fc874bf73ba958; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/org-ci-heavy/2026-11-09T05:48:36Z/a964b2f3fc74fba04a437b9ed04fa1bfae1b114f3a0dbff839fc874bf73ba958</id><title>Lab 08: a run on org-ci-heavy, record a964b2f3fc74, SAME</title><updated>2026-11-09T05:48:36Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-org-ci-heavy-20261109054836-a964b2f3fc74"/><summary>Can an agent fix a red build? The org-hosted runner, machine class org-ci-heavy, ran commit d934f799; record hash a964b2f3fc74fba04a437b9ed04fa1bfae1b114f3a0dbff839fc874bf73ba958; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-09T05:39:21Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on org-ci-heavy, record 04fa0d5ba8de, SAME</title><updated>2026-11-09T05:39:21Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261109053921-04fa0d5ba8de"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-09T05:39:20Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on ubuntu-latest, record 04fa0d5ba8de, SAME</title><updated>2026-11-09T05:39:20Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261109053920-04fa0d5ba8de"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-09T05:27:02Z/feb80a48b4596e6235b737288bcb0c87080354cb42411307a7acf9bc22de2827</id><title>Lab 05: a run on ubuntu-latest, record feb80a48b459, SAME</title><updated>2026-11-09T05:27:02Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261109052702-feb80a48b459"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit d934f799; record hash feb80a48b4596e6235b737288bcb0c87080354cb42411307a7acf9bc22de2827; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-09T05:26:39Z/feb80a48b4596e6235b737288bcb0c87080354cb42411307a7acf9bc22de2827</id><title>Lab 05: a run on org-ci-heavy, record feb80a48b459, SAME</title><updated>2026-11-09T05:26:39Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261109052639-feb80a48b459"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit d934f799; record hash feb80a48b4596e6235b737288bcb0c87080354cb42411307a7acf9bc22de2827; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-020-amendment-05</id><title>Amendment · EXP 020: MongoDB on its home ground: where does it fit Odin?</title><updated>2026-11-09T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/mongodb-home-ground-pre-registration.html"/><summary>S2&apos;s run record: the pause, all 13 discards and the report&apos;s hashes.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/ubuntu-latest/2026-11-08T05:56:21Z/9aaa4dcaaff096b77f33c729c87ab651ba2cd5c346ebedf39cf71fb6ce36e698</id><title>Lab 13: a run on ubuntu-latest, record 9aaa4dcaaff0, SAME</title><updated>2026-11-08T05:56:21Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-ubuntu-latest-20261108055621-9aaa4dcaaff0"/><summary>Does an agent company get better, measured blind? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 9aaa4dcaaff096b77f33c729c87ab651ba2cd5c346ebedf39cf71fb6ce36e698; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/org-ci-heavy/2026-11-08T05:55:50Z/9aaa4dcaaff096b77f33c729c87ab651ba2cd5c346ebedf39cf71fb6ce36e698</id><title>Lab 13: a run on org-ci-heavy, record 9aaa4dcaaff0, SAME</title><updated>2026-11-08T05:55:50Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-org-ci-heavy-20261108055550-9aaa4dcaaff0"/><summary>Does an agent company get better, measured blind? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 9aaa4dcaaff096b77f33c729c87ab651ba2cd5c346ebedf39cf71fb6ce36e698; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/ubuntu-latest/2026-11-08T05:49:02Z/6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05</id><title>Lab 08: a run on ubuntu-latest, record 6a7cf1bd96ba, SAME</title><updated>2026-11-08T05:49:02Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-ubuntu-latest-20261108054902-6a7cf1bd96ba"/><summary>Can an agent fix a red build? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/org-ci-heavy/2026-11-08T05:48:43Z/6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05</id><title>Lab 08: a run on org-ci-heavy, record 6a7cf1bd96ba, SAME</title><updated>2026-11-08T05:48:43Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-org-ci-heavy-20261108054843-6a7cf1bd96ba"/><summary>Can an agent fix a red build? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-08T05:38:22Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on ubuntu-latest, record 04fa0d5ba8de, SAME</title><updated>2026-11-08T05:38:22Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261108053822-04fa0d5ba8de"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-08T05:38:03Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on org-ci-heavy, record 04fa0d5ba8de, SAME</title><updated>2026-11-08T05:38:03Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261108053803-04fa0d5ba8de"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-08T05:26:01Z/58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f</id><title>Lab 05: a run on ubuntu-latest, record 58eff38051ea, SAME</title><updated>2026-11-08T05:26:01Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261108052601-58eff38051ea"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-08T05:25:44Z/58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f</id><title>Lab 05: a run on org-ci-heavy, record 58eff38051ea, SAME</title><updated>2026-11-08T05:25:44Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261108052544-58eff38051ea"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-021-amendment-02</id><title>Amendment · EXP 021: Kill the database mid-session, does anyone notice?</title><updated>2026-11-08T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/failover-results.html"/><summary>The pre-registered single-copy rule fired, so the compared PostgreSQL variants are strict; ten runner fixes are recorded.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/ubuntu-latest/2026-11-07T05:55:41Z/9aaa4dcaaff096b77f33c729c87ab651ba2cd5c346ebedf39cf71fb6ce36e698</id><title>Lab 13: a run on ubuntu-latest, record 9aaa4dcaaff0, SAME</title><updated>2026-11-07T05:55:41Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-ubuntu-latest-20261107055541-9aaa4dcaaff0"/><summary>Does an agent company get better, measured blind? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 9aaa4dcaaff096b77f33c729c87ab651ba2cd5c346ebedf39cf71fb6ce36e698; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/org-ci-heavy/2026-11-07T05:55:13Z/9aaa4dcaaff096b77f33c729c87ab651ba2cd5c346ebedf39cf71fb6ce36e698</id><title>Lab 13: a run on org-ci-heavy, record 9aaa4dcaaff0, SAME</title><updated>2026-11-07T05:55:13Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-org-ci-heavy-20261107055513-9aaa4dcaaff0"/><summary>Does an agent company get better, measured blind? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 9aaa4dcaaff096b77f33c729c87ab651ba2cd5c346ebedf39cf71fb6ce36e698; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/ubuntu-latest/2026-11-07T05:48:43Z/6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05</id><title>Lab 08: a run on ubuntu-latest, record 6a7cf1bd96ba, SAME</title><updated>2026-11-07T05:48:43Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-ubuntu-latest-20261107054843-6a7cf1bd96ba"/><summary>Can an agent fix a red build? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/org-ci-heavy/2026-11-07T05:48:19Z/6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05</id><title>Lab 08: a run on org-ci-heavy, record 6a7cf1bd96ba, SAME</title><updated>2026-11-07T05:48:19Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-org-ci-heavy-20261107054819-6a7cf1bd96ba"/><summary>Can an agent fix a red build? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-07T05:38:24Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on org-ci-heavy, record 04fa0d5ba8de, SAME</title><updated>2026-11-07T05:38:24Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261107053824-04fa0d5ba8de"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-07T05:38:08Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on ubuntu-latest, record 04fa0d5ba8de, SAME</title><updated>2026-11-07T05:38:08Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261107053808-04fa0d5ba8de"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-07T05:25:16Z/58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f</id><title>Lab 05: a run on ubuntu-latest, record 58eff38051ea, SAME</title><updated>2026-11-07T05:25:16Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261107052516-58eff38051ea"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-07T05:24:54Z/58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f</id><title>Lab 05: a run on org-ci-heavy, record 58eff38051ea, SAME</title><updated>2026-11-07T05:24:54Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261107052454-58eff38051ea"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/org-ci-heavy/2026-11-06T06:08:21Z/9aaa4dcaaff096b77f33c729c87ab651ba2cd5c346ebedf39cf71fb6ce36e698</id><title>Lab 13: a run on org-ci-heavy, record 9aaa4dcaaff0, SAME</title><updated>2026-11-06T06:08:21Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-org-ci-heavy-20261106060821-9aaa4dcaaff0"/><summary>Does an agent company get better, measured blind? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 9aaa4dcaaff096b77f33c729c87ab651ba2cd5c346ebedf39cf71fb6ce36e698; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/org-ci-heavy/2026-11-06T06:04:55Z/6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05</id><title>Lab 08: a run on org-ci-heavy, record 6a7cf1bd96ba, SAME</title><updated>2026-11-06T06:04:55Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-org-ci-heavy-20261106060455-6a7cf1bd96ba"/><summary>Can an agent fix a red build? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-06T06:04:30Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on org-ci-heavy, record 04fa0d5ba8de, SAME</title><updated>2026-11-06T06:04:30Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261106060430-04fa0d5ba8de"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/measured-blind/runs/ubuntu-latest/2026-11-06T05:56:23Z/9aaa4dcaaff096b77f33c729c87ab651ba2cd5c346ebedf39cf71fb6ce36e698</id><title>Lab 13: a run on ubuntu-latest, record 9aaa4dcaaff0, SAME</title><updated>2026-11-06T05:56:23Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-measured-blind-ubuntu-latest-20261106055623-9aaa4dcaaff0"/><summary>Does an agent company get better, measured blind? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 9aaa4dcaaff096b77f33c729c87ab651ba2cd5c346ebedf39cf71fb6ce36e698; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-06T05:56:00Z/58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f</id><title>Lab 05: a run on org-ci-heavy, record 58eff38051ea, SAME</title><updated>2026-11-06T05:56:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261106055600-58eff38051ea"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/ubuntu-latest/2026-11-06T05:48:26Z/6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05</id><title>Lab 08: a run on ubuntu-latest, record 6a7cf1bd96ba, SAME</title><updated>2026-11-06T05:48:26Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-ubuntu-latest-20261106054826-6a7cf1bd96ba"/><summary>Can an agent fix a red build? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-06T05:38:59Z/04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798</id><title>Lab 06: a run on ubuntu-latest, record 04fa0d5ba8de, SAME</title><updated>2026-11-06T05:38:59Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261106053859-04fa0d5ba8de"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 88779efe; record hash 04fa0d5ba8deadf3416591dfce2cfe08494299e4b84acaeae4aec67678058798; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-06T05:25:45Z/58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f</id><title>Lab 05: a run on ubuntu-latest, record 58eff38051ea, SAME</title><updated>2026-11-06T05:25:45Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261106052545-58eff38051ea"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/ubuntu-latest/2026-11-05T05:48:33Z/6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05</id><title>Lab 08: a run on ubuntu-latest, record 6a7cf1bd96ba, SAME</title><updated>2026-11-05T05:48:33Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-ubuntu-latest-20261105054833-6a7cf1bd96ba"/><summary>Can an agent fix a red build? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/org-ci-heavy/2026-11-05T05:48:06Z/6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05</id><title>Lab 08: a run on org-ci-heavy, record 6a7cf1bd96ba, SAME</title><updated>2026-11-05T05:48:06Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-org-ci-heavy-20261105054806-6a7cf1bd96ba"/><summary>Can an agent fix a red build? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-05T05:39:39Z/a43b8f02a2117b17e27ccdbbea5e6c9d1144faff04488233e3f2b3c43254df4e</id><title>Lab 06: a run on org-ci-heavy, record a43b8f02a211, SAME</title><updated>2026-11-05T05:39:39Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261105053939-a43b8f02a211"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit edc746aa; record hash a43b8f02a2117b17e27ccdbbea5e6c9d1144faff04488233e3f2b3c43254df4e; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-05T05:39:30Z/a43b8f02a2117b17e27ccdbbea5e6c9d1144faff04488233e3f2b3c43254df4e</id><title>Lab 06: a run on ubuntu-latest, record a43b8f02a211, SAME</title><updated>2026-11-05T05:39:30Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261105053930-a43b8f02a211"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit edc746aa; record hash a43b8f02a2117b17e27ccdbbea5e6c9d1144faff04488233e3f2b3c43254df4e; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-05T05:25:52Z/58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f</id><title>Lab 05: a run on ubuntu-latest, record 58eff38051ea, SAME</title><updated>2026-11-05T05:25:52Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261105052552-58eff38051ea"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-05T05:25:31Z/58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f</id><title>Lab 05: a run on org-ci-heavy, record 58eff38051ea, SAME</title><updated>2026-11-05T05:25:31Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261105052531-58eff38051ea"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/ubuntu-latest/2026-11-04T19:32:52Z/6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05</id><title>Lab 08: a run on ubuntu-latest, record 6a7cf1bd96ba, SAME</title><updated>2026-11-04T19:32:52Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-ubuntu-latest-20261104193252-6a7cf1bd96ba"/><summary>Can an agent fix a red build? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/red-build/runs/org-ci-heavy/2026-11-04T19:32:06Z/6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05</id><title>Lab 08: a run on org-ci-heavy, record 6a7cf1bd96ba, SAME</title><updated>2026-11-04T19:32:06Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-red-build-org-ci-heavy-20261104193206-6a7cf1bd96ba"/><summary>Can an agent fix a red build? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 6a7cf1bd96ba0bf3dfa931154b9862b0aa27253a67eb8106c71259651f18cb05; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-04T05:39:09Z/a43b8f02a2117b17e27ccdbbea5e6c9d1144faff04488233e3f2b3c43254df4e</id><title>Lab 06: a run on ubuntu-latest, record a43b8f02a211, SAME</title><updated>2026-11-04T05:39:09Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261104053909-a43b8f02a211"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit edc746aa; record hash a43b8f02a2117b17e27ccdbbea5e6c9d1144faff04488233e3f2b3c43254df4e; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-04T05:39:05Z/a43b8f02a2117b17e27ccdbbea5e6c9d1144faff04488233e3f2b3c43254df4e</id><title>Lab 06: a run on org-ci-heavy, record a43b8f02a211, SAME</title><updated>2026-11-04T05:39:05Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261104053905-a43b8f02a211"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit edc746aa; record hash a43b8f02a2117b17e27ccdbbea5e6c9d1144faff04488233e3f2b3c43254df4e; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-04T05:26:51Z/58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f</id><title>Lab 05: a run on ubuntu-latest, record 58eff38051ea, SAME</title><updated>2026-11-04T05:26:51Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261104052651-58eff38051ea"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-04T05:26:38Z/58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f</id><title>Lab 05: a run on org-ci-heavy, record 58eff38051ea, SAME</title><updated>2026-11-04T05:26:38Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261104052638-58eff38051ea"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-03T05:39:10Z/f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c</id><title>Lab 06: a run on ubuntu-latest, record f93aa888409b, SAME</title><updated>2026-11-03T05:39:10Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261103053910-f93aa888409b"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 0442d084; record hash f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-03T05:39:02Z/f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c</id><title>Lab 06: a run on org-ci-heavy, record f93aa888409b, SAME</title><updated>2026-11-03T05:39:02Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261103053902-f93aa888409b"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 0442d084; record hash f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-03T05:25:59Z/58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f</id><title>Lab 05: a run on ubuntu-latest, record 58eff38051ea, SAME</title><updated>2026-11-03T05:25:59Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261103052559-58eff38051ea"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit 58e84a9d; record hash 58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-03T05:25:45Z/58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f</id><title>Lab 05: a run on org-ci-heavy, record 58eff38051ea, SAME</title><updated>2026-11-03T05:25:45Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261103052545-58eff38051ea"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit 58e84a9d; record hash 58eff38051ea12e0387f20383f089d045d45bf7e0e8897003ec88fa0a114120f; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-02T05:39:32Z/f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c</id><title>Lab 06: a run on ubuntu-latest, record f93aa888409b, SAME</title><updated>2026-11-02T05:39:32Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261102053932-f93aa888409b"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 0442d084; record hash f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-02T05:39:03Z/f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c</id><title>Lab 06: a run on org-ci-heavy, record f93aa888409b, SAME</title><updated>2026-11-02T05:39:03Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261102053903-f93aa888409b"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 0442d084; record hash f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-02T05:26:09Z/8a49047402da6391d75bf72ff34ec54c2f5ee0b5f02900dfd2c89abb772cb065</id><title>Lab 05: a run on ubuntu-latest, record 8a49047402da, SAME</title><updated>2026-11-02T05:26:09Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261102052609-8a49047402da"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit 137f9822; record hash 8a49047402da6391d75bf72ff34ec54c2f5ee0b5f02900dfd2c89abb772cb065; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-02T05:25:55Z/8a49047402da6391d75bf72ff34ec54c2f5ee0b5f02900dfd2c89abb772cb065</id><title>Lab 05: a run on org-ci-heavy, record 8a49047402da, SAME</title><updated>2026-11-02T05:25:55Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261102052555-8a49047402da"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit 137f9822; record hash 8a49047402da6391d75bf72ff34ec54c2f5ee0b5f02900dfd2c89abb772cb065; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-11-01T05:38:46Z/f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c</id><title>Lab 06: a run on ubuntu-latest, record f93aa888409b, SAME</title><updated>2026-11-01T05:38:46Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261101053846-f93aa888409b"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 0442d084; record hash f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-11-01T05:38:41Z/f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c</id><title>Lab 06: a run on org-ci-heavy, record f93aa888409b, SAME</title><updated>2026-11-01T05:38:41Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261101053841-f93aa888409b"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 0442d084; record hash f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-11-01T05:26:07Z/a31ba363cfa79af9d453c666614e40c45f1cd8fcc1ea15989dba86a419c51f9c</id><title>Lab 05: a run on ubuntu-latest, record a31ba363cfa7, SAME</title><updated>2026-11-01T05:26:07Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261101052607-a31ba363cfa7"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit ccc31e2a; record hash a31ba363cfa79af9d453c666614e40c45f1cd8fcc1ea15989dba86a419c51f9c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-11-01T05:25:45Z/a31ba363cfa79af9d453c666614e40c45f1cd8fcc1ea15989dba86a419c51f9c</id><title>Lab 05: a run on org-ci-heavy, record a31ba363cfa7, SAME</title><updated>2026-11-01T05:25:45Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261101052545-a31ba363cfa7"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit ccc31e2a; record hash a31ba363cfa79af9d453c666614e40c45f1cd8fcc1ea15989dba86a419c51f9c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-10-31T05:38:38Z/f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c</id><title>Lab 06: a run on ubuntu-latest, record f93aa888409b, SAME</title><updated>2026-10-31T05:38:38Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261031053838-f93aa888409b"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 0442d084; record hash f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-10-31T05:38:11Z/f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c</id><title>Lab 06: a run on org-ci-heavy, record f93aa888409b, SAME</title><updated>2026-10-31T05:38:11Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261031053811-f93aa888409b"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 0442d084; record hash f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-10-31T05:25:45Z/a31ba363cfa79af9d453c666614e40c45f1cd8fcc1ea15989dba86a419c51f9c</id><title>Lab 05: a run on ubuntu-latest, record a31ba363cfa7, SAME</title><updated>2026-10-31T05:25:45Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261031052545-a31ba363cfa7"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit ccc31e2a; record hash a31ba363cfa79af9d453c666614e40c45f1cd8fcc1ea15989dba86a419c51f9c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-10-31T05:25:26Z/a31ba363cfa79af9d453c666614e40c45f1cd8fcc1ea15989dba86a419c51f9c</id><title>Lab 05: a run on org-ci-heavy, record a31ba363cfa7, SAME</title><updated>2026-10-31T05:25:26Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261031052526-a31ba363cfa7"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit ccc31e2a; record hash a31ba363cfa79af9d453c666614e40c45f1cd8fcc1ea15989dba86a419c51f9c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-10-30T05:40:01Z/f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c</id><title>Lab 06: a run on ubuntu-latest, record f93aa888409b, SAME</title><updated>2026-10-30T05:40:01Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261030054001-f93aa888409b"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 0442d084; record hash f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-10-30T05:39:42Z/f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c</id><title>Lab 06: a run on org-ci-heavy, record f93aa888409b, SAME</title><updated>2026-10-30T05:39:42Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261030053942-f93aa888409b"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 0442d084; record hash f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-10-30T05:27:45Z/a31ba363cfa79af9d453c666614e40c45f1cd8fcc1ea15989dba86a419c51f9c</id><title>Lab 05: a run on ubuntu-latest, record a31ba363cfa7, SAME</title><updated>2026-10-30T05:27:45Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261030052745-a31ba363cfa7"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit ccc31e2a; record hash a31ba363cfa79af9d453c666614e40c45f1cd8fcc1ea15989dba86a419c51f9c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-10-30T05:27:29Z/a31ba363cfa79af9d453c666614e40c45f1cd8fcc1ea15989dba86a419c51f9c</id><title>Lab 05: a run on org-ci-heavy, record a31ba363cfa7, SAME</title><updated>2026-10-30T05:27:29Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261030052729-a31ba363cfa7"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit ccc31e2a; record hash a31ba363cfa79af9d453c666614e40c45f1cd8fcc1ea15989dba86a419c51f9c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/ubuntu-latest/2026-10-29T22:47:03Z/f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c</id><title>Lab 06: a run on ubuntu-latest, record f93aa888409b, SAME</title><updated>2026-10-29T22:47:03Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-ubuntu-latest-20261029224703-f93aa888409b"/><summary>Which model should review your code? The github-hosted runner, machine class ubuntu-latest, ran commit 0442d084; record hash f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/review/runs/org-ci-heavy/2026-10-29T22:46:52Z/f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c</id><title>Lab 06: a run on org-ci-heavy, record f93aa888409b, SAME</title><updated>2026-10-29T22:46:52Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-review-org-ci-heavy-20261029224652-f93aa888409b"/><summary>Which model should review your code? The org-hosted runner, machine class org-ci-heavy, ran commit 0442d084; record hash f93aa888409b042edfac057f12ad82ff4d2c75e33d07f8daf93547eb4e3e262c; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-10-29T05:28:06Z/47c9ab93279dae96f8423aabc985beda6492f2e498da82e055d9ba9fa44aacf8</id><title>Lab 05: a run on ubuntu-latest, record 47c9ab93279d, SAME</title><updated>2026-10-29T05:28:06Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261029052806-47c9ab93279d"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit aa35e6db; record hash 47c9ab93279dae96f8423aabc985beda6492f2e498da82e055d9ba9fa44aacf8; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-10-29T05:27:41Z/47c9ab93279dae96f8423aabc985beda6492f2e498da82e055d9ba9fa44aacf8</id><title>Lab 05: a run on org-ci-heavy, record 47c9ab93279d, SAME</title><updated>2026-10-29T05:27:41Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261029052741-47c9ab93279d"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit aa35e6db; record hash 47c9ab93279dae96f8423aabc985beda6492f2e498da82e055d9ba9fa44aacf8; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/ubuntu-latest/2026-10-28T07:25:27Z/47c9ab93279dae96f8423aabc985beda6492f2e498da82e055d9ba9fa44aacf8</id><title>Lab 05: a run on ubuntu-latest, record 47c9ab93279d, SAME</title><updated>2026-10-28T07:25:27Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-ubuntu-latest-20261028072527-47c9ab93279d"/><summary>Can an agent go rogue? The github-hosted runner, machine class ubuntu-latest, ran commit aa35e6db; record hash 47c9ab93279dae96f8423aabc985beda6492f2e498da82e055d9ba9fa44aacf8; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/org-ci-heavy/2026-10-28T07:25:13Z/47c9ab93279dae96f8423aabc985beda6492f2e498da82e055d9ba9fa44aacf8</id><title>Lab 05: a run on org-ci-heavy, record 47c9ab93279d, SAME</title><updated>2026-10-28T07:25:13Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-org-ci-heavy-20261028072513-47c9ab93279d"/><summary>Can an agent go rogue? The org-hosted runner, machine class org-ci-heavy, ran commit aa35e6db; record hash 47c9ab93279dae96f8423aabc985beda6492f2e498da82e055d9ba9fa44aacf8; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-020-amendment-04</id><title>Amendment · EXP 020: MongoDB on its home ground: where does it fit Odin?</title><updated>2026-10-28T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/mongodb-home-ground-pre-registration.html"/><summary>Both advocates settle nine readings of the S2 rules before the verdict; both veto item 8, so the report checks the disturbance windows itself.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/labs/rogue/runs/mac/2026-10-27T11:25:33Z/47c9ab93279dae96f8423aabc985beda6492f2e498da82e055d9ba9fa44aacf8</id><title>Lab 05: a run on mac, record 47c9ab93279d, SAME</title><updated>2026-10-27T11:25:33Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/labs/runs/#run-rogue-mac-20261027112533-47c9ab93279d"/><summary>Can an agent go rogue? The laptop runner, machine class mac, ran commit aa35e6db; record hash 47c9ab93279dae96f8423aabc985beda6492f2e498da82e055d9ba9fa44aacf8; the check printed SAME with 0 breaches. This is a check of the committed record; it does not establish a new experiment or independent validation.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-020-amendment-01</id><title>Amendment · EXP 020: MongoDB on its home ground: where does it fit Odin?</title><updated>2026-10-25T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/mongodb-home-ground-pre-registration.html"/><summary>S2&apos;s runner changes, the rule for which cells count, and the nine discarded cells are published.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-020-amendment-02</id><title>Amendment · EXP 020: MongoDB on its home ground: where does it fit Odin?</title><updated>2026-10-25T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/mongodb-home-ground-pre-registration.html"/><summary>S2 compares MongoDB with one PostgreSQL variant per scale, chosen once.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-020-amendment-03</id><title>Amendment · EXP 020: MongoDB on its home ground: where does it fit Odin?</title><updated>2026-10-25T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/mongodb-home-ground-pre-registration.html"/><summary>Both advocates accept the S2 pairing rule; the decision paragraph is rewritten as they required.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-020-preregistration</id><title>Pre-registered · EXP 020: MongoDB on its home ground: where does it fit Odin?</title><updated>2026-10-25T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/mongodb-home-ground-pre-registration.html"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-021-amendment-01</id><title>Amendment · EXP 021: Kill the database mid-session, does anyone notice?</title><updated>2026-10-25T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/failover-results.html"/><summary>The runner protocol and the design draft are published.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-021-preregistration</id><title>Pre-registered · EXP 021: Kill the database mid-session, does anyone notice?</title><updated>2026-10-25T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/failover-results.html"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-mongodb-ai-context</id><title>Field note: When a cached yes answers the opposite question</title><updated>2026-10-25T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/mongodb-ai-context.html"/><summary>We changed the rules behind saved AI answers and reworded questions so yes meant the opposite. Versions kept old answers out, but similar wording still caused wrong replies.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-mongodb-ai-ecosystem</id><title>Field note: MongoDB beyond JSON: agents, memory and models</title><updated>2026-10-24T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/mongodb-ai-ecosystem.html"/><summary>We tested MongoDB for agents and memory with real Voyage embeddings. Both databases found the same nearest results; tools can repeat after a crash and cached answers can outlive a policy change.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-015-preregistration</id><title>Pre-registered · EXP 015: Does Odin&apos;s cost estimate match what the brokers bill?</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/data/telemetry-truth/preregistration.json"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-016-amendment-01</id><title>Amendment · EXP 016: Does the same model cost the same through every broker?</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/broker-same-model-results.html"/><summary>Read the full record, conditions and limitations at the linked page. This entry announces a publication; it makes no result claim.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-016-preregistration</id><title>Pre-registered · EXP 016: Does the same model cost the same through every broker?</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/broker-same-model-results.html"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-016-results</id><title>Results · EXP 016: Does the same model cost the same through every broker?</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/broker-same-model-results.html"/><summary>Measured on 720 identical calls: two brokers billed exactly list price for Claude Sonnet 5.5 and GPT-4o mini, the third 10% above it on both, a spread of 1.10 against a pre-registered bar of 1.05. No broker served a different model.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-017-preregistration</id><title>Pre-registered · EXP 017: What does a defect caught in review cost, model by model?</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/review-frontier-results.html"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-017-results</id><title>Results · EXP 017: What does a defect caught in review cost, model by model?</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/review-frontier-results.html"/><summary>Read the full record, conditions and limitations at the linked page. This entry announces a publication; it makes no result claim.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-018-preregistration</id><title>Pre-registered · EXP 018: Does routing by our own rulebook beat one fixed model on real tasks?</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/router-arm-results.html"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-018-results</id><title>Results · EXP 018: Does routing by our own rulebook beat one fixed model on real tasks?</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/router-arm-results.html"/><summary>Read the full record, conditions and limitations at the linked page. This entry announces a publication; it makes no result claim.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-019-preregistration</id><title>Pre-registered · EXP 019: Routing by our own rulebook, measured again: three fresh runs of the routed arm</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/router-arm-fresh-results.html"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-019-results</id><title>Results · EXP 019: Routing by our own rulebook, measured again: three fresh runs of the routed arm</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/router-arm-fresh-results.html"/><summary>Pre-registered verdict: routing beats always-Opus, within a wide replicate allowance, and does not beat always-Luna.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-broker-same-model-results</id><title>Field note: EXP 016: Does the same model cost the same through every broker?</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/broker-same-model-results.html"/><summary>Almost. Measured on 720 identical calls: two brokers billed exactly list price for Claude Sonnet 5.5 and GPT-4o mini, the third 10% above it on both, a spread of 1.10 against a pre-registered bar of 1.05. No broker served a different model.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-review-frontier-results</id><title>Field note: EXP 017: What does a defect caught in review cost, model by model?</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/review-frontier-results.html"/><summary>Read the full record, conditions and limitations at the linked page. This entry announces a publication; it makes no result claim.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-router-arm-fresh-results</id><title>Field note: EXP 019: Routing by our own rulebook, measured again</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/router-arm-fresh-results.html"/><summary>Read the full record, conditions and limitations at the linked page. This entry announces a publication; it makes no result claim.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-router-arm-results</id><title>Field note: EXP 018: Does routing by our own rulebook beat one fixed model on real tasks?</title><updated>2026-10-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/router-arm-results.html"/><summary>Read the full record, conditions and limitations at the linked page. This entry announces a publication; it makes no result claim.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-012-results</id><title>Results · EXP 012: Can a reviewer talk itself out of its false alarms?</title><updated>2026-10-22T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/self-verify-results.html"/><summary>No. Asked to name a concrete failing input for every finding or drop it, MiniMax M3.1 Flash kept 249 of 269 of its own findings. False alarms stayed at 20 and 22 of 30 clean changes, against a pre-registered bar of 2.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-self-verify-results</id><title>Field note: EXP 012: Can a reviewer talk itself out of its false alarms?</title><updated>2026-10-22T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/self-verify-results.html"/><summary>No. Asked to name a concrete failing input for every finding or drop it, MiniMax M3.1 Flash kept 249 of 269 of its own findings. False alarms stayed at 20 and 22 of 30 clean changes, against a pre-registered bar of 2.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-012-amendment-01</id><title>Amendment · EXP 012: Can a reviewer talk itself out of its false alarms?</title><updated>2026-10-21T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/self-verify-results.html"/><summary>Read the full record, conditions and limitations at the linked page. This entry announces a publication; it makes no result claim.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-012-preregistration</id><title>Pre-registered · EXP 012: Can a reviewer talk itself out of its false alarms?</title><updated>2026-10-21T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/self-verify-results.html"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-flash-on-the-review-bench</id><title>Field note: A free model on our review bench</title><updated>2026-10-21T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/flash-on-the-review-bench.html"/><summary>Read the full record, conditions and limitations at the linked page. This entry announces a publication; it makes no result claim.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-free-model-own-bug-history</id><title>Field note: Can a free model find real bugs in our private code?</title><updated>2026-10-21T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/free-model-own-bug-history.html"/><summary>On bugs in our private fix history that can be found from the code alone, MiniMax M3.1 Flash&apos;s first guess landed within two lines of the fix 63% of the time, against 23% for our current default. Only about 1 in 10 of the small fixes we sampled could be found that way.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-mongodb-primitive-fit</id><title>Field note: Where would MongoDB earn a place in Odin?</title><updated>2026-10-19T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/mongodb-primitive-fit.html"/><summary>99,000 requests and a separate 19,800-request repeat found two advantages over monolithic JSONB, but none over the stronger PostgreSQL alternatives, for these workloads on one machine. All seventeen outcomes and the durability amendment are retained.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-mongodb-temporal-document-fit</id><title>Field note: Does flexible workflow data need MongoDB?</title><updated>2026-10-19T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/mongodb-temporal-document-fit.html"/><summary>Both stores passed 20,000 operations, 42 correctness checks and ten evolution suites behind Temporal activities. For this workload, a MongoDB advantage remains unresolved and flexible JSON alone gave no demonstrated reason to switch.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-005-erratum-01</id><title>Erratum · EXP 005: Jev as a fast gate</title><updated>2026-10-11T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/jev-as-a-fast-gate.html"/><summary>Erratum 01: the DOI link for Newcombe (1998) in the pre-registration does not resolve.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-011-preregistration</id><title>Pre-registered · EXP 011: Where does a model reviewer earn its keep?</title><updated>2026-10-11T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/where-a-model-reviewer-earns-its-keep.html"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-009-results</id><title>Results · EXP 009: Unload leaves no trace</title><updated>2026-10-10T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/composable-harness-results.html"/><summary>P1 refuted: 1 of 400 H1 trials diverged from R0 (95% Wilson upper bound 1.40%). P2 supported: 0 silent-inert and 0 false-inactive kernel configurations of 120. All 5 validity gates passed, so the record is informative.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-008-amendment-02</id><title>Amendment · EXP 008: Hand over the cache, not the text</title><updated>2026-10-09T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/latent-handoff-pre-registration.html"/><summary>Amendment 02: the zero-prefill counter fix and a fresh counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-007-results</id><title>Results · EXP 007: Which of your gate&apos;s rules need a model at all?</title><updated>2026-10-08T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/which-rules-need-a-model.html#results"/><summary>Pre-registered; measured 7 Oct 2026: the premise is refuted.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-007-amendment-01</id><title>Amendment · EXP 007: Which of your gate&apos;s rules need a model at all?</title><updated>2026-10-07T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/which-rules-need-a-model.html#results"/><summary>Amendment 01: one re-call for the translator calls the client&apos;s spend limit refused.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-009-preregistration</id><title>Pre-registered · EXP 009: Unload leaves no trace</title><updated>2026-10-07T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/composable-harness-results.html"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-007-preregistration</id><title>Pre-registered · EXP 007: Which of your gate&apos;s rules need a model at all?</title><updated>2026-10-03T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/which-rules-need-a-model.html#results"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-008-amendment-01</id><title>Amendment · EXP 008: Hand over the cache, not the text</title><updated>2026-10-02T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/latent-handoff-pre-registration.html"/><summary>Amendment 01: the confirmatory analysis script.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-008-preregistration</id><title>Pre-registered · EXP 008: Hand over the cache, not the text</title><updated>2026-10-02T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/latent-handoff-pre-registration.html"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-006-amendment-01</id><title>Amendment · EXP 006: nina reviews the change</title><updated>2026-10-01T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/nina-reviews-the-change.html#amendment-01"/><summary>Amendment 01: a digest is not a restricted term, a compound git call counts, and the client&apos;s built-in reads are disclosed.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-006-results</id><title>Results · EXP 006: nina reviews the change</title><updated>2026-10-01T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/nina-reviews-the-change.html#amendment-01"/><summary>Pre-registered; measured 01 OCT 2026: bar met, nina in the spotlight. Amended 01 Oct 2026, before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-006-preregistration</id><title>Pre-registered · EXP 006: nina reviews the change</title><updated>2026-09-30T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/nina-reviews-the-change.html#amendment-01"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-005-results</id><title>Results · EXP 005: Jev as a fast gate</title><updated>2026-09-29T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/jev-as-a-fast-gate.html"/><summary>Pre-registered; measured 29 Sep 2026: the claim is refuted.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-005-amendment-01</id><title>Amendment · EXP 005: Jev as a fast gate</title><updated>2026-09-28T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/jev-as-a-fast-gate.html"/><summary>Amendment 01: nina 0.34.0 and a pre-registered spotlight bar.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-005-amendment-02</id><title>Amendment · EXP 005: Jev as a fast gate</title><updated>2026-09-28T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/jev-as-a-fast-gate.html"/><summary>Amendment 02: the reviewer&apos;s file and git tools are fenced to its workspace, in two layers.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-005-preregistration</id><title>Pre-registered · EXP 005: Jev as a fast gate</title><updated>2026-09-28T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/jev-as-a-fast-gate.html"/><summary>The question and how it will be scored were published before any counted run.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/tool-laya-browser</id><title>Tool: Laya in your browser</title><updated>2026-09-25T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/projects/laya-browser/"/><summary>Run Laya, an open-weight typed-decision model, in your browser after a model download. Read its probabilities for a typed question. Three hand-written cases checked parity with the reference; they did not measure accuracy or calibration. Inference runs on your device without an account or API key.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-004-measured</id><title>Measured · EXP 004: System-1 decisions without lock-in: Jev vs open-weight Laya</title><updated>2026-09-24T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/system-1-decisions-without-lock-in.html"/><summary>Measured on one laptop, 48 hand-written rows.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/exp-001-003-recorded</id><title>Recorded · EXP 001–003: Three blueprint recipes</title><updated>2026-09-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/experiments/#bench"/><summary>Each recipe applies one unchanged architectural rule to a conforming tree and to a deliberately drifted tree; the engine scored every conforming tree GREEN and every drifted tree RED.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/tool-ci-witness</id><title>Tool: CI Witness</title><updated>2026-09-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/projects/ci-witness/"/><summary>Turn an environment-sensitive failure into an inspectable comparison: the failing assertion, a controlled reproduction and a correction that keeps the agreed checks.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/tool-migration-witness</id><title>Tool: Migration Witness</title><updated>2026-09-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/projects/migration-witness/"/><summary>Make the acceptance decision explicit: compare the old and new component against examples you approve, with intentional changes recorded separately.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/tool-test-witness</id><title>Tool: Test Witness</title><updated>2026-09-23T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/projects/test-witness/"/><summary>A passing suite can miss a broken calculation. Inspect whether tests expose specified defects before you rely on their green result.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-a-passing-pipeline</id><title>Field note: A passing pipeline was not yet a verified release.</title><updated>2026-09-07T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/a-passing-pipeline.html"/><summary>What a missing coverage report taught us about making “done” explainable.</summary></entry><entry><id>tag:odin-labs-ai.github.io,2026:odin-rnd/updates/note-why-open-the-floor</id><title>Field note: Why open the factory floor?</title><updated>2026-09-07T00:00:00Z</updated><link rel="alternate" type="text/html" href="https://odin-labs-ai.github.io/odin-rnd/journal/why-open-the-floor.html"/><summary>Tools, experiments and a place to show work while it is still becoming something.</summary></entry></feed>
