tooluniverse-self-review

Generate the success criteria for a task or question, then review work against them. Given a task, goal, or open-ended question, decompose it into scenarios, evaluation perspectives, and fine-grained weighted YES/NO criteria using the Recursive Expansion Tree (RET) method; if work is supplied, score it criterion-by-criterion and surface what is missing or could be better. Use when asked to self-review or check your own work, judge whether a task is done well or completely, build a definition-of-done or completeness checklist, create an evaluation rubric or grading criteria, score or grade answers to a question, set up an LLM-as-judge rubric, or when the user mentions self-review, completeness check, success criteria, evaluation criteria, scoring rubric, Qworld, or the RET algorithm.

Quality

88%

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

Quality

Content

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A highly actionable, clearly sequenced procedural skill with strong validation checkpoints, but it is monolithic and longer than it needs to be due to restated principles and inline detail that could be split out. Progressive disclosure is the weakest area.

Suggestions

Dedupe the 'Core Principles', step bodies, and 'Key Constraints' sections — state each rule once (e.g., 'no fixed dimension lists', binary criteria, mandatory dedup) and cross-reference rather than restating, to tighten conciseness.

Extract the detailed per-step output-format specs and scoring standard into a reference file (e.g. references/RET_FORMAT.md), keeping SKILL.md as a lean overview that links one level deep, to improve progressive disclosure.

Move the BibTeX citation block to a references/CITATION.md file or a short footnote, so the operational body stays focused on the workflow.

Dimension	Reasoning	Score
Conciseness	Largely procedural and necessary, but ~350 lines with redundancy — 'No fixed dimension lists', binary criteria, and deduplication are restated across 'Core Principles', the step bodies, and 'Key Constraints', and the Algorithm Overview re-explains the steps that follow. Score 3 is barred by the repetition; score 1 is too harsh given the content is mostly task-specific, not generic concept padding.	2 / 3
Actionability	Provides fully concrete guidance: exact expansion round counts (3/4/3), a 10-step + Phase B checklist, named output fields, and a specific scoring standard (positive 1-10, negative -1 to -10) — copy-paste ready, matching the score-3 anchor.	3 / 3
Workflow Clarity	Clear sequenced workflow with a copyable progress checklist and explicit review/consolidation checkpoints (Steps 5, 8, 9, 10) that act as feedback loops (expand -> review -> fix), matching the score-3 anchor.	3 / 3
Progressive Disclosure	Single monolithic file with no bundle files or external references; sections are well-organized (ruling out score 1) but content that could be split — detailed output-format specs, the citation block, expansion-round mechanics — is kept inline, fitting the score-2 anchor better than score 3 which expects one-level-deep references.	2 / 3
	Total	10 / 12 Passed

Description

100%

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 concretely states what the skill does and gives an explicit, keyword-rich 'Use when...' clause. It is specific, complete, and unlikely to conflict with other skills.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions such as 'Generate the success criteria', 'decompose it into scenarios, evaluation perspectives, and fine-grained weighted YES/NO criteria', 'score it criterion-by-criterion', and 'build a definition-of-done or completeness checklist', matching the score-3 anchor.	3 / 3
Completeness	Explicitly answers both what ('Generate the success criteria... then review work against them') and when ('Use when asked to self-review or check your own work...'), with an explicit trigger clause, matching the score-3 anchor.	3 / 3
Trigger Term Quality	Surfaces natural terms a user would say — 'self-review', 'check your own work', 'definition-of-done', 'completeness checklist', 'evaluation rubric', 'grading criteria', 'score or grade answers', 'LLM-as-judge', 'scoring rubric', plus the distinctive 'Qworld' and 'RET algorithm' — giving broad coverage.	3 / 3
Distinctiveness Conflict Risk	Occupies a clear niche (task-specific success-criteria derivation and criterion-by-criterion work review) with distinctive triggers (Qworld, RET, definition-of-done) unlikely to fire for unrelated skills.	3 / 3
	Total	12 / 12 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: mims-harvard/ToolUniverse
Path: plugins/tooluniverse/skills/tooluniverse-self-review/SKILL.md
Commit: 089eb8e

Reviewed: about 7 hours 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.