fix-pr-issues

Use when addressing Basic Memory pull request feedback, failed checks, or Codex review blockers, then re-running the latest-head review gate.

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

92%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

An exemplary instruction-only skill body: a tight gather/fix/recheck/reply loop with explicit validation gates, a ledger-driven feedback cycle, and zero wasted tokens. The only gap is that a few steps (comment collection, pushing) lack the concrete commands that the gather step's `gh pr view` demonstrates.

Suggestions

Add the specific commands for feedback collection (e.g. `gh pr view --comments`, `gh run view --log <job-id>`) to make the Gather phase fully copy-paste executable like its `gh pr view` step.

State the push command explicitly (e.g. `git push` or `git push origin <head>`) instead of the bare directive 'Push the branch'.

Dimension	Reasoning	Score
Conciseness	The ~44-line body is lean and directive with no padding or explanations of concepts Claude already knows ('Keep diffs narrow and preserve unrelated user changes', 'This skill never merges a PR' each earn their place). Every token is load-bearing, matching anchor 5.	5 / 5
Actionability	Concrete, executable guidance is present: the exact `gh pr view --json number,url,headRefOid,mergeStateStatus,statusCheckRollup` command, `git commit -s`, and precise conditions like 'zero unresolved, non-outdated Codex threads'. It is below anchor 5 because feedback collection and push are described only at a high level with no commands ('Push the branch', 'PR comments and review summaries').	4 / 5
Workflow Clarity	A clear four-phase sequence (Gather, Fix, Push And Recheck, Reply) with explicit validation checkpoints ('watch them to completion', 'require the current-head approval signal') and an explicit error-recovery feedback loop ('If new feedback appears, add it to the ledger and repeat the loop'), matching anchor 5.	5 / 5
Progressive Disclosure	The skill is under 50 lines, needs no external reference files (the only bundle file is an icon), and is organized into clear headed sections; per the rubric's simple-skill guideline, this scores 5. The `pr-review-loop` pointer is a named sibling skill, not a nested reference file.	5 / 5
	Total	19 / 20 Passed

Description

78%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A tight, well-scoped description with an explicit trigger clause and domain-specific terms that make it highly distinguishable. Its main limitation is that the capability statement leans on the generic verb 'addressing' and embeds the 'what' inside the 'when' clause rather than enumerating concrete actions.

Suggestions

State the 'what' as an explicit action list before the 'Use when' clause, e.g. 'Fixes PR review feedback, failed checks, and Codex review blockers, then re-runs the latest-head review gate.'

Add natural trigger synonyms users would say, such as 'CI failures', 'review comments', or 'unresolved review threads', to broaden trigger term coverage.

Consider mentioning the reply/resolve-thread outcome so the description's action coverage is comprehensive rather than implicitly carried by the body.

Dimension	Reasoning	Score
Specificity	The description names the domain and several specific inputs ('pull request feedback, failed checks, or Codex review blockers') plus a concrete closing action ('re-running the latest-head review gate'), matching anchor 4's 'several specific actions; minor gaps'. It is not anchor 5 because the governing verb 'addressing' is generic and the action list is not comprehensive (no mention of replying to threads, pushing, or verification runs).	4 / 5
Completeness	It has an explicit 'Use when...' clause with concrete triggers, and a 'what' (fix blockers, then re-run the review gate), so both are present. It stops short of anchor 5 because the 'what' is folded into the when-clause rather than stated as an explicit, self-contained action list.	4 / 5
Trigger Term Quality	'pull request feedback', 'failed checks', and 'Codex review blockers' are natural phrases a contributor on this repo would actually say, giving good keyword coverage. It is below anchor 5 because common variations and synonyms like 'CI failures', 'review comments', or 'fix the build' are missing.	4 / 5
Distinctiveness Conflict Risk	Scoping to 'Basic Memory' PRs, 'Codex review blockers', and the 'latest-head review gate' carves out a clear niche with distinct triggers and minimal conflict risk with generic PR, git, or code-review skills.	5 / 5
	Total	17 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: basicmachines-co/basic-memory
Path: .agents/skills/fix-pr-issues/SKILL.md
Commit: 194afe1

Reviewed: 7 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.