CtrlK
BlogDocsLog inGet started
Tessl Logo

jbaruch/coding-policy

General-purpose coding policy for Baruch's AI agents

74

Quality

93%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Medium

Suggest reviewing before use

Overview
Quality
Evals
Security
Files

skill-review-gate.mdskills/release/references/

Skill Review Gate

Detail of the rules/context-artifacts.md skill-review gate, moved out of the always-loaded rule (#642). It binds as rule content: the rule keeps the gate's directives and each carve-out's trigger line, and requires reading this file before relying on them. Each section carries only the moved detail; the directives and the carve-out's trigger, applies-when and reset lines stay in the rule.

Credit-Outage Review Carve-Out

  • The skill-review action's credit-outage: skip input publishes the affected skill unreviewed rather than blocking every skill-changing merge until credits return
  • Preconditions (all required):
    1. Opt-in explicit — the consumer sets credit-outage: skip; default fail preserves the gate. Classification is the action's decision contract — see .github/actions/skill-review/review-skills.sh run_reviews
    2. Fail-safe — only the out-of-credits signature skips; every other non-zero exit still hard-fails the publish (below-threshold score, auth error, tooling bug)
    3. Each unreviewed publish is flagged — a ::warning:: plus a $GITHUB_STEP_SUMMARY note names every skipped skill, surfaced on the unreviewed-skills action output
    4. The gate self-heals — every publish re-attempts review; a credit top-up or the monthly reset restores it
  • The forbidden review-step bypass in rules/context-artifacts.md Disagreeing With the Reviewer remains forbidden

CI Wiring

  • Wire review into CI through .github/actions/skill-review/action.yml (consumers uses: jbaruch/coding-policy/.github/actions/skill-review@<ref>), a changed-skills loop, never static per-skill steps
  • Which skills a run reviews, and the fallback when no diff base is available, are the action's contract — see the description block at the top of .github/actions/skill-review/action.yml and .github/actions/skill-review/review-skills.sh
  • Inputs: threshold, skills-dir, base-ref, workspace, credit-outage — defaults and meaning in the action's inputs: block
  • Output: unreviewed-skills, naming every skill a credit-outage skip published unreviewed
  • Failure: a below-threshold score, or any review failure the credit-outage input does not cover, fails the step and blocks the publish
  • Rubric verifies: frontmatter validity, execution-mode preamble matching the skill's shape (sequential workflow or action router per rules/skill-authoring.md), flat step numbering, typed Skill() calls, silence-rule compliance, channel-appropriate formatting

Fix Loop

  • When you disagree with the reviewer's conclusions, run tessl review fix <skill> locally
  • Back up SKILL.md and any reference files before invoking the fix loop
  • The fix loop is a diagnostic signal, not a patch
  • Diff the applied changes against the backup
  • Keep the genuinely-improving moves (tighter triggers, less prose, better Skill() typing)
  • Reject the over-aggressive cuts
  • Re-run tessl review run until the gate passes
  • Curate the fixes manually with judgment

skills

README.md

tile.json