General-purpose coding policy for Baruch's AI agents
73
91%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
Low
Low-risk findings worth noting
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.
skill-review action's credit-outage: skip input publishes the affected skill unreviewed rather than blocking every skill-changing merge until credits returncredit-outage: skip; default fail preserves the gate. Classification is the action's decision contract — see .github/actions/skill-review/review-skills.sh run_reviews::warning:: plus a $GITHUB_STEP_SUMMARY note names every skipped skill, surfaced on the unreviewed-skills action outputrules/context-artifacts.md Disagreeing With the Reviewer remains forbidden.github/actions/skill-review/action.yml (consumers uses: jbaruch/coding-policy/.github/actions/skill-review@<ref>), a changed-skills loop, never static per-skill stepsdescription block at the top of .github/actions/skill-review/action.yml and .github/actions/skill-review/review-skills.shthreshold, skills-dir, base-ref, workspace, credit-outage — defaults and meaning in the action's inputs: blockunreviewed-skills, naming every skill a credit-outage skip published unreviewedrules/skill-authoring.md), flat step numbering, typed Skill() calls, silence-rule compliance, channel-appropriate formattingtessl review fix <skill> locallySKILL.md and any reference files before invoking the fix loopSkill() typing)tessl review run until the gate passes.tessl-plugin
hooks
rules
skills
adopt-fork-pr
herdr-foreman
classify
foreman
references
templates
tests
herdr-standup
migrate-to-plugin
onboard-repo
release
references
tests