fix: target Level 3 in nightly reference enrichment#299
Merged
Conversation
The cron was using --min-level 2 which only found Level 0-1 agents, missing the 22 Level 0 and 5 Level 2 agents that need enrichment. Changed to --min-level 3 so every agent below Level 3 (Deep) is a candidate. The goal is continuous improvement, not just backfilling. Also removed hardcoded level<2 filter in select-enrichment-targets.py since the audit's --min-level flag already handles filtering.
notque
added a commit
that referenced
this pull request
Apr 11, 2026
The audit flags anything below Level 3, but the prompt said "enrich to Level 2+" and "skip if already Level 2+". Agents at Level 2 were flagged forever and skipped every run — never processed. c5f89eb (#299) upgraded the audit but left the prompt text unchanged; observed 2026-04-11 with react-native-engineer.
3 tasks
notque
added a commit
that referenced
this pull request
Apr 11, 2026
…371) The audit flags anything below Level 3, but the prompt said "enrich to Level 2+" and "skip if already Level 2+". Agents at Level 2 were flagged forever and skipped every run — never processed. c5f89eb (#299) upgraded the audit but left the prompt text unchanged; observed 2026-04-11 with react-native-engineer.
notque
added a commit
that referenced
this pull request
Apr 11, 2026
…rget Four stale threshold references in skills/reference-enrichment/SKILL.md that were missed when #299 moved the audit to --min-level 3: - Line 30 top-level description: "Level 0-1 to Level 2-3" → "Level 0-2 to Level 3+" - Line 102 Phase 4 VALIDATE goal: "Level 2+ depth" → "Level 3+ depth" - Line 108 verification instruction: "level field is 2 or 3 ... If still Level 1" → "level field is 3 ... If still below Level 3" - Line 165 failure case heading: "Phase 4 Tier 1 still Level 1" → "Phase 4 Tier 1 still below Level 3" Companion to #371 which fixed the runtime prompt text. SKILL.md is documentation about the skill and not read by the cron, so the urgency was lower, but leaving these inconsistent would confuse anyone reading the skill and wondering why the audit threshold does not match the stated validation criteria.
3 tasks
notque
added a commit
that referenced
this pull request
Apr 11, 2026
…rget (#372) Four stale threshold references in skills/reference-enrichment/SKILL.md that were missed when #299 moved the audit to --min-level 3: - Line 30 top-level description: "Level 0-1 to Level 2-3" → "Level 0-2 to Level 3+" - Line 102 Phase 4 VALIDATE goal: "Level 2+ depth" → "Level 3+ depth" - Line 108 verification instruction: "level field is 2 or 3 ... If still Level 1" → "level field is 3 ... If still below Level 3" - Line 165 failure case heading: "Phase 4 Tier 1 still Level 1" → "Phase 4 Tier 1 still below Level 3" Companion to #371 which fixed the runtime prompt text. SKILL.md is documentation about the skill and not read by the cron, so the urgency was lower, but leaving these inconsistent would confuse anyone reading the skill and wondering why the audit threshold does not match the stated validation criteria.
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.
Summary
The enrichment cron was using
--min-level 2which only found Level 0-1 agents — missing 27 agents that need enrichment (22 at Level 0, 5 at Level 2).Changed to
--min-level 3so every agent below Level 3 (Deep) is a candidate. The goal is continuous depth improvement toward Level 3, not just backfilling missing references.Changes
scripts/reference-enrichment-cron.sh:--min-level 2→--min-level 3scripts/select-enrichment-targets.py: removed hardcodedlevel < 2filter (the audit's--min-levelflag already handles filtering)Test plan
bash -n scripts/reference-enrichment-cron.sh— syntax validruff check+ruff format --check— clean