auto-review-loop

Autonomous multi-round research review loop. Repeatedly reviews using a secondary Codex agent, implements fixes, and re-reviews until positive assessment or max rounds reached. Use when user says "auto review loop", "review until it passes", or wants autonomous iterative improvement.

Quality

73%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/skills-codex/auto-review-loop/SKILL.md

Quality

Content

63%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.

A thorough, actionable multi-phase review loop with strong sequencing and validation gates, undermined by verbatim duplication and inline material that should be split out. Tightening repetition and externalizing the scope-limits/test specs would lift conciseness and progressive disclosure.

Suggestions

De-duplicate the SCOPE LIMITS block: reference review-scope-limits.md once instead of pasting the full block in both the medium and round-2+ prompts.

Move the 'Acquittal Gate Test Specifications' section into a separate bundled reference file (e.g. references/acquittal-tests.md) and link to it, keeping SKILL.md focused on execution.

Consolidate the reviewer-memory rules, which currently appear both under '## Claude-Aligned Reviewer Memory and Debate' and '#### Phase B.5', into a single authoritative location.

Dimension	Reasoning	Score
Conciseness	Mostly efficient operational prose, but the ~20-line SCOPE LIMITS block is duplicated verbatim in the medium and round-2+ prompts, reviewer-memory mechanics are described twice, and the appended Acquittal Gate Test Specifications inflate the file without aiding execution.	3 / 5
Actionability	Provides copy-paste spawn_agent/send_input message templates, concrete JSON state schemas, and exact commands (e.g. '/render-html "review-stage/AUTO_REVIEW.md" --no-review --state review-stage/REVIEW_STATE.json'); templates carry placeholders, so just shy of fully executable anchor 5.	4 / 5
Workflow Clarity	Phases A–E plus Initialization and Termination are clearly sequenced with explicit checkpoints (Stop-Evaluation Gate, training-quality check, Debate Protocol, append-only acquittal); minor gaps such as reviewer-failure handling are only briefly noted.	4 / 5
Progressive Disclosure	Sections are well-labeled and external references (../shared-references/*.md) are signaled with links, but no bundle files exist in the skill, the SCOPE LIMITS content is inlined+duplicated rather than referenced, and the test-spec block is inline material that belongs elsewhere.	3 / 5
	Total	14 / 20 Passed

Description

83%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 strong, third-person description that clearly states capability and trigger conditions with concrete natural phrases. Minor headroom remains on action specificity and trigger synonym coverage.

Dimension	Reasoning	Score
Specificity	Names the domain and three concrete actions — 'reviews using a secondary Codex agent', 'implements fixes', and 're-reviews' — plus loop termination logic; not quite the comprehensive multi-action coverage of the anchor-5 example.	4 / 5
Completeness	Explicitly answers both what (autonomous review→fix→re-review loop) and when ('Use when user says...') with concrete trigger phrases, matching the anchor-5 example.	5 / 5
Trigger Term Quality	Quotes three natural trigger phrases ('auto review loop', 'review until it passes', 'autonomous iterative improvement') a user would plausibly say, though synonym/extension coverage is not exhaustive.	4 / 5
Distinctiveness Conflict Risk	Clear niche (autonomous research-review loop via a secondary Codex agent) with specific triggers and minimal overlap risk; 'autonomous iterative improvement' is slightly broad, keeping it just below anchor 5.	4 / 5
	Total	17 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (501 lines); consider splitting into references/ and linking	Warning
relative_links	Relative link issues: 4 suspicious	Warning

	Total	14 / 16 Passed

Repository: wanshuiyin/Auto-claude-code-research-in-sleep
Path: skills/skills-codex/auto-review-loop/SKILL.md
Commit: 0472e53

Reviewed: 5 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.