review

Code review with semantic diffs, expert routing, and auto-TaskCreate. Triggers on: code review, review changes, check code, review PR, security audit.

1.34x

Quality

—

Does it follow best practices?

Impact

85%

1.34x

Average score across 3 eval scenarios

Securityby

Advisory

Suggest reviewing before use

Quality

Content

65%

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

The content is highly actionable with concrete executable commands and a clear six-step workflow, but it is verbose due to duplicated step descriptions and lacks a validation feedback loop on the destructive --fix path. Progressive disclosure is weak: the single referenced file is missing and most material is inline.

Suggestions

Collapse the ASCII architecture diagram and the Execution Steps section into one sequenced walkthrough to remove the duplicated six-step restatement.

Add a re-review validation loop to the --fix path (e.g., re-run the review after applying edits and only mark the TaskCreate item complete when the issue no longer appears).

Create the referenced framework-checks.md (or remove the dangling reference) and move the framework-specific patterns and the full review output template into it to slim the body and fix the broken reference.

Dimension	Reasoning	Score
Conciseness	The body is token-heavy: the ASCII architecture diagram and the Execution Steps section restate the same six steps, and the full mock review template (SQL injection, useEffect examples) runs long. It avoids explaining concepts Claude already knows, so it is not a 1, but the duplication and padding keep it from being lean.	2 / 3
Actionability	It provides copy-paste-ready executable bash (git diff --cached --name-only, gh pr diff, command -v delta fallbacks), a concrete file-pattern routing table, and literal TaskCreate/TaskUpdate invocations, matching fully executable guidance.	3 / 3
Workflow Clarity	The six-step sequence is clearly ordered with some checkpoints (staged-change check, tool-availability guards, --fix confirmation), but the destructive --fix path applies edits and marks tasks complete without a re-review validation loop, capping it at 2.	2 / 3
Progressive Disclosure	Sections are well organized, but ~250 lines are monolithic in SKILL.md and the one signaled reference (framework-checks.md) does not exist in any bundle directory; content that should be split (framework checks, full output template) remains inline.	2 / 3
	Total	9 / 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.

The description is concise, concrete, and answers both what the skill does and when to invoke it with natural trigger terms. It is well-differentiated and avoids fluff.

Dimension	Reasoning	Score
Specificity	"Code review with semantic diffs, expert routing, and auto-TaskCreate" lists three concrete, distinct capabilities rather than vague abstractions, matching the anchor for multiple specific concrete actions.	3 / 3
Completeness	It states what the skill does (semantic diffs, expert routing, auto-TaskCreate) and explicitly when to use it via an equivalent "Triggers on:" clause, satisfying both the what and the when.	3 / 3
Trigger Term Quality	"Triggers on: code review, review changes, check code, review PR, security audit" are natural phrases a user would actually say, giving good coverage of common variations.	3 / 3
Distinctiveness Conflict Risk	The code-review niche is framed by differentiating features (semantic diffs, expert routing, auto-TaskCreate) and triggers are scoped to review activity, making conflict with unrelated skills unlikely.	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: NeverSight/skills_feed
Commit: 26d37ce

Reviewed: 4 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.