chore(deps): bump Mattraks/delete-workflow-runs from 998b89248266543a3c7827da0a73cf49d0131222 to b3018382ca039b53d238908238bd35d1fb14f8ee - #1
Open
dependabot[bot] wants to merge 2 commits into
Conversation
Bumps [Mattraks/delete-workflow-runs](https://github.com/mattraks/delete-workflow-runs) from 998b89248266543a3c7827da0a73cf49d0131222 to b3018382ca039b53d238908238bd35d1fb14f8ee. - [Release notes](https://github.com/mattraks/delete-workflow-runs/releases) - [Commits](Mattraks/delete-workflow-runs@998b892...b301838) --- updated-dependencies: - dependency-name: Mattraks/delete-workflow-runs dependency-version: b3018382ca039b53d238908238bd35d1fb14f8ee dependency-type: direct:production ... Signed-off-by: dependabot[bot] <support@github.com>
dependabot
Bot
added
dependencies
mevBlaze
added a commit
that referenced
this pull request
Jul 1, 2026
…ssion The app submits OTP email to our own Stalwart MTA over the host-internal hop (same box). Stalwart presents a self-signed cert there; SMTP_TLS_INSECURE=1 lets the submission proceed (rejectUnauthorized:false). This is NOT the public delivery hop — Stalwart→Gmail uses the real cert + DKIM (mail._domainkey.pocketbits.in). Env-gated; no behavior change until set. Pre-staged so the dev-stub→live flip is instant once gilfoyle provisions recovery@pocketbits.in + its password. Real PAN/phone KYC already hash-loaded into user_identity (13,884 users, staging dropped, wall-proven) — Verifier prerequisite #1 of 2 done. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016jtEm8s9mxR2GmPScq13At
mevBlaze
added a commit
that referenced
this pull request
Jul 1, 2026
Upgrades oracle_health from a simple :HealthWarning node reader (which would return HEALTHY even with 4,299 isolated cortex dots) to a full brain self-check grounded in live Cypher counts. New sections: - SCALE: Observations, CortexEvents, total edges, edges/node ratio - CONNECTIVITY: % nodes with ≥1 edge, island counts, cross-layer SIMILAR - EDGE BREAKDOWN: typed reasoning vs SIMILAR vs Hebbian by type - DECISION-RANK CANARY: live ranker test on OBS-axxis-20260430-1519 (COIN NAME IS PIED PIPER) for query "Pied Piper HBO trademark" Must rank #1; PASS/FAIL derived from real ranker output - HONEST STATUS: GREEN only when canary PASS + islands within threshold + edges/node ≥ 5; never a bare green Adds _compute_brain_vitals() with 60s TTL cache, _run_canary() that mirrors the REST /search pipeline (fresh vectors, over-fetch, raw channel exclusion, pre-warm call). force=true arg bypasses cache. Verified live on VPS: all numbers match direct cypher-shell counts. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016jtEm8s9mxR2GmPScq13At
mevBlaze
added a commit
that referenced
this pull request
Jul 1, 2026
…bove-R67 PASS criteria _run_canary() previously created fresh TreeVectorService+TreeSearch and used rerank=True — giving different results than the real oracle_query code path (singleton, rerank=False). The lock appeared at rank #1 (reranker-boosted) in the canary but rank #3 in real queries — a false-positive PASS signal. Fix: - Use get_search() singleton (same as handle_query / oracle_query) - rerank=False, include_synthesis=True — exact handle_query signature - Over-fetch top_k*5, post-filter channel:raw in Python — mirrors handle_query - Handle _vectors=None gracefully (SKIP status instead of ERROR) PASS criteria updated (rigid #1 was a brittle false-fail): - Lock (OBS-axxis-20260430-1519) must be in TOP-3 - Lock must rank above R67 ruling (OBS-axxis-20260429-1324/1328) Newer high-relevance Blaze decisions legitimately rank above the lock (same ×1.8 policy boost but higher recency/Hebbian) — they belong above it. Canary verified live: rank #3 score 0.6644, R67 obs absent from top-10 (correctly demoted far below), PASS → oracle_health GREEN HEALTHY. Score-drop investigation (0.9126 earlier this session → 0.6644 now): NOT densification. NOT PageRank dilution (PageRank not in blend_scores). Root cause: two new Blaze obs created TODAY (OBS-piper-20260630-006/008) have recency≈1.0 (created hours ago) vs lock's recency≈0.0 (61 days, lambda=0.01/hour → e^-14.6≈0) and Hebbian 160/70 vs lock's 12. These raise max_rrf and max_heb normalization denominators, reducing the lock's normalized blend score. Expected correct behavior — not a regression. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016jtEm8s9mxR2GmPScq13At
mevBlaze
added a commit
that referenced
this pull request
Jul 1, 2026
…e 9, user-facing path) Previous fix used rerank=False (oracle_query agent path) — wrong for the canary. Anti-rot Rule 9: user-facing /search defaults rerank=True. The canary tests whether the brain correctly surfaces locked decisions to users — that's the user-facing path. The original bug was fresh TreeVectorService instances (cold vector index, unstable candidates) not the rerank=True choice. Fix keeps the singleton (warm, stable) and restores rerank=True (TEI cross-encoder, port 4085). Verified live: lock OBS-axxis-20260430-1519 at rank #1 score 1.7892, R67 ruling OBS-axxis-20260429-1328 at #3 (below lock), OBS-axxis-20260429-1324 absent — oracle_health GREEN HEALTHY. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016jtEm8s9mxR2GmPScq13At
mevBlaze
added a commit
that referenced
this pull request
Jul 4, 2026
Front door + 3 tiers by load-trigger (LAW always-on / CONTEXT load-on-demand / DIARY-STATE) + the duplication map (autonomy/oracle-protocol/TTS stated in 3 places) + roadmap (state.md history bloat is #1). Blaze: 'better way of managing rules?' Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016jtEm8s9mxR2GmPScq13At
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Bumps Mattraks/delete-workflow-runs from 998b89248266543a3c7827da0a73cf49d0131222 to b3018382ca039b53d238908238bd35d1fb14f8ee.
Commits
b301838Upgrade to Node 24bd2822cUpdate deprecated@octokit/restlibrary call methods4c5df81Update log message for retaining runs5bf9a1dBug Fixes63b223fSupports daily retention keep a minimum number of runs per day86d29a7Update deprecated@octokit/restlibrary call methodsab48244version v2.1.0feeb820Undo previous changes.c61e04eAdded another try/catch blockf93a693Added another try/catch blockDependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting
@dependabot rebase.Dependabot commands and options
You can trigger Dependabot actions by commenting on this PR:
@dependabot rebasewill rebase this PR@dependabot recreatewill recreate this PR, overwriting any edits that have been made to it@dependabot show <dependency name> ignore conditionswill show all of the ignore conditions of the specified dependency@dependabot ignore this major versionwill close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this minor versionwill close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this dependencywill close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)