auto-review-loop-llm

Autonomous research review loop using any OpenAI-compatible LLM API. Configure via llm-chat MCP server or environment variables. Trigger with "auto review loop llm" or "llm review".

Quality

60%

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/auto-review-loop-llm/SKILL.md

Quality

Content

57%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 actionable with concrete MCP/curl templates and a clear phased loop, but it is moderately verbose and lacks validation checkpoints on the fix-implementation step. Progressive disclosure is adequate but inlines content that would benefit from separate reference files.

Suggestions

Add an explicit validation step in Phase C (e.g. run tests/lints or re-check claims before re-review) to create a fix → validate → re-review feedback loop.

Move the provider table and the Round 2+ prompt template into separate reference files and link to them from the body to reduce inlined bulk.

Trim editorial commentary (e.g. the stale-verdict aside in Constants) and de-duplicate the MCP/curl blocks to improve token efficiency.

Dimension	Reasoning	Score
Conciseness	Mostly efficient with concrete templates, but contains editorial padding (e.g. the stale-verdict commentary in Constants) and duplicated MCP/curl blocks that could be tightened.	3 / 5
Actionability	Provides concrete MCP tool invocations, curl commands, JSON state files, and prompt templates that are mostly executable, with only minor gaps such as placeholder shell variables.	4 / 5
Workflow Clarity	The phased A–E loop with recovery and an explicit STOP condition is clearly sequenced, but Phase C (implement fixes) lacks any validation/verification checkpoint before re-review; for iterative/batch research changes this caps the score at 3.	3 / 5
Progressive Disclosure	Single-file content with reasonable section headers and one-level-deep shared-references links, but sizable inlined material (provider table, full prompt templates) that could live in separate reference files leaves organization only moderately strong.	3 / 5
	Total	13 / 20 Passed

Description

62%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 conveys a clear purpose and includes explicit trigger phrases, but its action coverage is narrow and trigger synonyms are sparse. It is distinct enough from other skills with low conflict risk.

Suggestions

Add more natural trigger variations and synonyms (e.g. "review my paper", "improve research draft", "auto-review") to broaden keyword coverage.

Expand the 'what' to name the concrete review actions (score, rank weaknesses, specify minimum fixes) rather than only the loop phases.

Add a 'Use when...' clause naming scenarios (e.g. preparing a manuscript for a top venue) to make the 'when' more explicit.

Dimension	Reasoning	Score
Specificity	Names the domain ("Autonomous research review loop") and concrete actions ("review → implement fixes → re-review"), but coverage is limited to those steps without naming what the review targets or what fixes entail.	3 / 5
Completeness	Has a clear "what" (autonomous research improvement loop) and an explicit "when" via the "Trigger with..." clause, but the trigger guidance is narrow and could be more explicit about scenarios.	4 / 5
Trigger Term Quality	Includes explicit trigger phrases ("auto review loop llm", "llm review") but offers few natural synonyms or variations a user might spontaneously say, so coverage is partial.	3 / 5
Distinctiveness Conflict Risk	The LLM-driven autonomous review-loop niche is fairly distinct with specific trigger terms, with only minor overlap risk against generic review skills.	4 / 5
	Total	14 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
relative_links	Relative link issues: 3 suspicious	Warning

	Total	15 / 16 Passed

Repository: wanshuiyin/Auto-claude-code-research-in-sleep
Path: skills/skills-codex/auto-review-loop-llm/SKILL.md
Commit: f4f20f9

Reviewed: 1 day 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.