aris-auto-review-loop

Autonomous multi-round research review loop. Repeatedly reviews via Codex 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

69%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

Fix and improve this skill with Tessl

tessl review fix ./skills/aris-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.

The body is highly detailed and actionable with a clear phased workflow, but it is verbose and monolithic — most reference material (prompt templates, log formats) is inlined rather than split into bundled files. Tightening and extracting templates would meaningfully improve it.

Suggestions

Extract the large verbatim reviewer prompt templates (medium/hard/nightmare) and log/markdown templates into files under ./references/ and link to them, improving progressive disclosure and token efficiency.

Tighten redundant explanatory paragraphs (e.g. re-explaining the loop mechanics across Context, Workflow intro, and Phase descriptions) to assume more of Claude's competence.

Add an explicit verification checkpoint in Phase C/D confirming implemented fixes actually executed/succeeded before Phase A re-review, to strengthen the feedback loop.

Dimension	Reasoning	Score
Conciseness	The skill is thorough but very long; large verbatim prompt templates for medium/hard/nightmare plus full markdown templates for logs and rebuttals repeat material Claude could derive, and several explanatory paragraphs restate workflow mechanics already implied by the structure.	3 / 5
Actionability	Provides concrete, copy-paste-ready MCP calls, codex exec bash snippets, and JSON/markdown templates for each phase; minor gaps are placeholder tokens ([Round N], [paste...]) that still require assembly.	4 / 5
Workflow Clarity	Multi-phase loop (A-F) is explicitly sequenced with checkpoints (stop condition, validation, state persistence, resume logic); the batch/destructive review loop has feedback loops, though some validation steps (e.g. confirming fixes actually ran before re-review) are implicit.	4 / 5
Progressive Disclosure	Content is monolithic within SKILL.md (no bundle files in references/scripts/assets, all detail inlined) and large verbatim prompt templates that could live in separate reference files are inlined; section headers help but the bulk is not split out.	3 / 5
	Total	14 / 20 Passed

Description

75%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 strong: it names the core loop, cites the external reviewer mechanism, and provides explicit "Use when..." trigger phrases. It is slightly verbose and could tighten the action list and add a few more natural synonyms.

Suggestions

Trim redundant phrasing (e.g. "Repeatedly reviews" + "re-reviews" overlap) to make the action list more comprehensive and concise.

Add a couple of natural synonyms to trigger terms such as "iterate on review" or "fix my paper" to broaden keyword coverage.

Make the "when" clause more explicit about the research/paper-review context to reduce overlap with generic review skills.

Dimension	Reasoning	Score
Specificity	Concrete actions are named ("repeatedly reviews", "implements fixes", "re-reviews") but verbs are somewhat minimal and not comprehensive; lacks enumeration of supporting actions like state persistence or memory tracking.	4 / 5
Completeness	Clearly states what (review via Codex MCP, implement fixes, re-review) and when (explicit "Use when..." with trigger phrases); both present, though the "when" could be more explicit about the research/paper context.	4 / 5
Trigger Term Quality	Includes natural user phrases ("auto review loop", "review until it passes", "wants autonomous iterative improvement") but misses common synonyms like "keep reviewing", "iterate on review", or "fix my paper".	4 / 5
Distinctiveness Conflict Risk	The autonomous loop + Codex MCP external-reviewer niche is fairly distinct from one-shot review skills; minor overlap risk with general review skills, and "autonomous iterative improvement" is slightly broad.	4 / 5
	Total	16 / 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
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	14 / 16 Passed

Repository: OpenLAIR/dr-claw
Path: skills/aris-auto-review-loop/SKILL.md
Commit: 4b35756

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.