claude-botandClaude Fable 5.1 25f09e9b52 ci(pick-model): newest among comparable candidates, policy in code ...

First live pick after #893 chose claude-sonnet-4-5 at 0.23 over
claude-sonnet-4-6: same-family models split the probability for a
mid-weight task, and "prefer newer" is a policy, not a judgment. New
scripts/ci/choose-model.sh applies it to the distribution: among
candidates within CHOOSE_MODEL_MARGIN (0.7) of the best probability the
most recently released wins; a clear best stays; NONE yields nothing.
The picker logs the top probabilities. Four policy tests added.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>

2026-09-21 09:59:40 +02:00

..

ai-eval

feat(scripts): capability eval to compare models on our own workloads

2026-09-01 17:21:38 +02:00

ci

ci(pick-model): newest among comparable candidates, policy in code

2026-09-21 09:59:40 +02:00

embed-eval

feat(scripts): retrieval eval for embedding models on German documents

2026-09-02 07:28:31 +02:00

paperless-eval

fix(paperless-eval): say whether the probe is loading or just not scheduled

2026-09-04 16:03:23 +02:00

ai-eval.sh

feat(scripts): capability eval to compare models on our own workloads

2026-09-01 17:21:38 +02:00

authentik-disable-2fa.sh

fix: Scope OBC health check to each app's own bucket

2026-04-19 00:39:52 +02:00

embed-eval.sh

feat(scripts): retrieval eval for embedding models on German documents

2026-09-02 07:28:31 +02:00

import-cluster-access.sh

feat(tooling): Hetzner throwaway site and rebuild-harness hardening

2026-09-06 20:47:45 +02:00

migrate-authentik.py

fix: Handle empty API responses in migration script, fix bootstrap

2026-04-13 09:14:11 +02:00

restore-authentik-db.sh

fix: Simplify restore script using mc in PG pod

2026-04-13 16:22:51 +02:00

trust-letsencrypt-staging.sh

chore: remove orphan manifests, harden CI fail-paths, drop stale CLAUDE.md

2026-04-29 18:37:04 +02:00