auto-review-loop

Autonomous multi-round research review loop. Repeatedly reviews using Claude Code via claude-review MCP, 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

78%

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

Fix and improve this skill with Tessl

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

Quality

Content

62%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.

The workflow is exceptionally well-sequenced with strong validation gates and executable MCP call templates, but the body is heavily verbose with duplicated prompt blocks and repeated assurance banners, and progressive disclosure is only partial since large prompt/scope-limit text is inlined rather than moved to reference files.

Suggestions

Deduplicate the two SCOPE LIMITS blocks: keep them only in review-scope-limits.md and reference it from both prompt templates instead of pasting the full text twice.

Move the large reviewer prompt templates and the acquittal/test specifications into reference files (e.g. prompts.md, acquittal-tests.md) and summarize them in SKILL.md to cut the ~30KB body down toward a lean overview.

Consolidate the repeated cross-family / acceptance_status assurance (stated in the frontmatter quote, the H1 banner, and the constants) into a single concise note.

Dimension	Reasoning	Score
Conciseness	At ~30KB / 500+ lines the body is markedly verbose: scope-limit blocks are duplicated verbatim in two prompt templates, reviewer-memory prose is restated multiple times, and header banners repeat the cross-family assurance, adding padding Claude does not need.	2 / 5
Actionability	Provides concrete, copy-pasteable MCP call blocks with explicit tools, prompt bodies, and polling instructions, plus exact JSONL/JSON schemas; a few placeholders (e.g. '<path>') keep it from being fully turn-key.	4 / 5
Workflow Clarity	The Initialization → Phase A–E loop is explicitly sequenced with a clear STOP CONDITION gate, acquittal write rules, and validation/checkpoint feedback at each transition; the append-only integrity rules and resume logic are unambiguous.	5 / 5
Progressive Disclosure	No bundle files exist and most detail is inlined; references to shared files (review-scope-limits.md, output-* protocols, review-tracing.md) are signaled but live outside the bundle and one (review-scope-limits.md) is duplicated inline rather than split, leaving organization only partially clean.	3 / 5
	Total	14 / 20 Passed

Description

95%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.

The description is concrete, third-person, and clearly answers both what the skill does and when to use it, with natural trigger phrases. It is well-distinguished from sibling skills and avoids vague fluff.

Dimension	Reasoning	Score
Specificity	Names the domain (autonomous review loop) and several concrete actions — 'Repeatedly reviews', 'implements fixes', and 're-reviews' — but coverage is thin relative to the many phases in the body, leaving minor gaps.	4 / 5
Completeness	Explicitly states what it does ('Repeatedly reviews...implements fixes, and re-reviews until positive assessment or max rounds reached') and when to use it with concrete trigger phrases, satisfying both requirements.	5 / 5
Trigger Term Quality	Provides natural phrases users would say verbatim — 'auto review loop', 'review until it passes' — plus the behavioral trigger 'wants autonomous iterative improvement', giving comprehensive natural coverage.	5 / 5
Distinctiveness Conflict Risk	The niche is distinctive — an autonomous iterative Claude-Code review loop routed through claude-review MCP — with idiosyncratic trigger phrases unlikely to collide with other skills.	5 / 5
	Total	19 / 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 (517 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-claude-review/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.