Self review

Skill tykisgod/quick-question/skills/self-review

Review changes from the most recent interaction (skills, configs, settings, and other lightweight changes) for quality and consistency.From its SKILL.md

Install

npx -y skills add tykisgod/quick-question --skill self-review

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

10 stars10 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
runs commandsInstructs the agent to run 2 commands, including `source "${CLAUDE_PLUGIN_ROOT}/scripts/platform/detect.sh"` and 1 more.

SKILL.md

3.1 KB, 699 tokens by cl100k_base, as published. Nobody here has run it

Respond in the user's preferred language (detect from their recent messages, or fall back to the language setting in CLAUDE.md).

Review changes from the most recent interaction via a subagent review loop. Automatically loops until no critical issues remain or 5 rounds are completed.

For Unity changes: when verifying that a runtime/scene change actually behaves as intended, use tykit live queries (unity_query / get-field / console) instead of re-reading the diff. See shared/tykit-first.md.

Steps

1. Collect changed files

Look back at the files changed in the most recent interaction and list them (file paths + one-line summary of each change).

2-6. Automated Review Loop

Loop automatically. Terminates when:

No [Critical] issues in the review
5 rounds completed
No new critical issues in two consecutive rounds

Output === Round N/5 === at the start of each round.

a. Dispatch review subagent

Dispatch a subagent (subagent_type: "general-purpose", model: "opus") with a prompt containing:

The list of changed files and a summary of what was changed
The full current content of each changed file (read them yourself first, then paste into the prompt — the subagent cannot read your session history)
Review checklist:
- Logical correctness — are references correct, are step numbers sequential, do file paths exist
- Consistency — does the style match existing content in this repo
- Omissions — are there related files that should have been updated but were missed
- Redundancy — unnecessary blank lines, duplicate content
- Naming — are renamed symbols/paths consistent everywhere
Required output format: classify each finding as [Critical], [Moderate], or [Suggestion] with file path, line number, and explanation

b. Fix confirmed issues

For each [Critical] issue: fix immediately
For each [Moderate] issue: fix at discretion
[Suggestion] items: note but do not fix unless trivial

c. Determine whether to continue

If this round had [Critical] issues fixed → start next round (back to a)
If no [Critical] issues → output "Review passed" and proceed to cleanup
If two consecutive rounds had no new critical issues → output "Review passed" and proceed to cleanup
If 5 rounds completed → output final status and proceed to cleanup

7. Clean up

Output a brief review conclusion, then clear the skill change marker:

source "${CLAUDE_PLUGIN_ROOT}/scripts/platform/detect.sh"
rm -f "$QQ_TEMP_DIR/claude-skill-modified-marker-$PPID"

Handoff

After the review loop ends, recommend the next step:

Review passed, no issues → "All clean. Ready to /qq:commit-push?"
Issues were found and fixed → "Fixed N issues across M rounds. Want to /qq:commit-push?"
5 rounds exhausted with remaining issues → "Some issues remain after 5 rounds. Please review manually before committing."

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most review quality skills give in 699 tokens

Counted across 1,273 of the 2,403 authors here whose files we hold, read 2026-09-06

Ask one question at a timein 63 of 1273, across 62 files
Provide a recommended answer for each questionin 47 of 1273, across 45 files
Rank findings by severityin 44 of 1273
Use parameterized queries for database accessin 38 of 1273, across 20 files
Validate all user input with schemasin 33 of 1273, across 15 files
Store secrets in environment variablesin 32 of 1273, across 14 files
Explore the codebase to answer questionsin 31 of 1273, across 29 files
Store tokens in httpOnly cookiesin 30 of 1273, across 12 files
Implement rate limiting on API endpointsin 30 of 1273, across 12 files
Sanitize user-provided HTMLin 29 of 1273, across 11 files
Return generic error messages to usersin 28 of 1273, across 10 files
Cite file and line for every findingin 28 of 1273, across 25 files

Said here and by no other author read

Recommend the next step after the loop ends
List changed files and summaries
Fix moderate issues at discretion
Output round status at start of each loop
Terminate if no critical issues remain
Clear skill change marker after review

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.