CtrlK
BlogDocsLog inGet started
Tessl Logo

uinaf/review-gang

Review existing code, diffs, branches, or pull requests by spawning mandatory concern-specific reviewer subagents, then synthesize a ship-it / needs-review / blocked verdict.

85

1.13x
Quality

89%

Does it follow best practices?

Impact

77%

1.13x

Average score across 4 eval scenarios

SecuritybySnyk

Passed

No known issues

Overview
Quality
Evals
Security
Files

Quality

Content

77%

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

An actionable, well-sequenced review skill with concrete prompts, commands, and an explicit output format. It would tighten with less repetition across the Contract and Codex sections, and the referenced bundle files must actually be provided for progressive disclosure to land.

Suggestions

Consolidate the subagent-discovery guidance into one place; it currently appears in the Contract, 'Before You Start' step 4, and the 'For Codex specifically' paragraph.

Provide the missing references/ directory with reviewer-selection.md and reviewing.md, or inline the essential reviewer-selection criteria so the referenced paths resolve.

Trim overlapping Contract bullets that restate workflow steps (e.g. scope definition, stale-evidence handling) to reduce redundancy.

DimensionReasoningScore

Conciseness

Mostly efficient and free of basic-concept padding, but subagent discovery is restated across the Contract, 'Before You Start' step 4, and the dedicated 'For Codex specifically' paragraph, and several Contract bullets overlap the workflow steps.

2 / 3

Actionability

Provides a copy-paste subagent prompt template, concrete git commands ('git diff --stat <base>...HEAD', 'pnpm test path/to/spec'), an explicit labeled output schema, and a worked example — all directly executable.

3 / 3

Workflow Clarity

Clear sequenced process (numbered 'Before You Start' 1–6 plus a four-step Workflow) with explicit validation checkpoints: 'Block when missing context or proof prevents an honest verdict' and refreshing stale CI/branch heads before judging.

3 / 3

Progressive Disclosure

Sections are well-organized and references are clearly signaled one level deep, but the body points to references/reviewer-selection.md and references/reviewing.md while no references/ bundle directory exists, so the navigation path leads to missing files.

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 description: concrete actions, an explicit 'Use when' trigger with natural phrasing, and a clearly bounded niche that distinguishes it from generic review skills. Third-person voice is used throughout, so no specificity penalty applies.

DimensionReasoningScore

Specificity

Lists multiple concrete actions — 'audit existing code, diffs, branches, or pull requests', 'spawning mandatory concern-specific reviewer subagents', 'synthesizing their evidence into a ship decision', and 'Produces a ship it / needs review / blocked verdict'.

3 / 3

Completeness

Answers both what (independent multi-reviewer audit synthesizing a ship decision) and when, with an explicit 'Use when triaging PR risk, deciding whether someone else's change is safe to ship, or following up after runtime proof' clause.

3 / 3

Trigger Term Quality

Natural user phrases like 'triaging PR risk', 'whether someone else's change is safe to ship', and 'following up after runtime proof' alongside 'diffs, branches, or pull requests' give good coverage of terms users would actually say.

3 / 3

Distinctiveness Conflict Risk

Clear niche of mandatory concern-specific reviewer subagents producing a ship verdict, with distinct triggers and an explicit 'Do not use to self-check a change you just authored' exclusion that reduces conflict risk.

3 / 3

Total

12

/

12

Passed

Validation

87%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 5 missing

Warning

referenced_paths_exist

Referenced path issues: 10 missing

Warning

Total

14

/

16

Passed

Reviewed

Table of Contents