CtrlK
BlogDocsLog inGet started
Tessl Logo

cursor-review

Run a Cursor Composer review as an advisory second opinion while keeping the primary review verdict on the host brain. Use when user invokes cursor:review, asks Cursor to review, or wants composer to sanity-check a diff. Cursor never owns APPROVE/REQUEST_CHANGES.

75

Quality

93%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

The canonical home for this skill is cursor-review in Chachamaru127/claude-code-harness

SKILL.md
Quality
Evals
Security

Quality

Content

87%

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

The content is lean, executable, and well-organized for a single-purpose skill, scoring top marks on conciseness, actionability, and structure. Its one weakness is workflow_clarity: the flow has no explicit validation or feedback checkpoint for Cursor output or host-verified findings.

Suggestions

Add an explicit checkpoint after step 3, e.g. 'If Cursor returns no findings or errors, note it and proceed to the host review' so the flow has a validation/feedback loop that can raise workflow_clarity to anchor 3.

Verify the referenced scripts exist in the bundle — scripts/resolve-impl-backend.sh, scripts/model-routing.sh, and scripts/cursor-companion.sh are cited but no scripts/ directory is present, which undermines the actionability of those commands.

Clarify in step 4 how the host should validate Cursor-originated findings (e.g. reproduce or confirm before marking them advisory) to give the primary-review step a concrete verification sub-step.

DimensionReasoningScore

Conciseness

The body is lean — a one-line intro, two quick examples, terse rules, and executable bash — with no padding explaining what Cursor or git diff is, matching the 'every token earns its place' anchor rather than anchor 2's unnecessary-explanation case.

3 / 3

Actionability

It gives complete, copy-paste-ready bash for scope parsing, git diff capture, and invoking cursor-companion, plus resolver/model-routing commands — matching anchor 3's fully-executable bar rather than the pseudocode/incomplete anchor 2.

3 / 3

Workflow Clarity

The four steps are clearly sequenced (resolve root → scope → ask Cursor → host verdict) but there is no explicit validation/feedback checkpoint, e.g. handling empty Cursor output or verifying findings before the host verdict, which fits anchor 2's 'checkpoints missing or implicit' rather than anchor 3's explicit validation steps.

2 / 3

Progressive Disclosure

The skill is single-purpose, self-contained, and cleanly sectioned (Quick Reference, Rules, Flow, Output) with no nested references, so under the simple-skills note the well-organized structure earns anchor 3 even without external bundle files.

3 / 3

Total

11

/

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 concrete, well-triggered, complete, and sharply scoped to an advisory Cursor-only role, hitting the top anchor on every dimension. It names explicit natural triggers and a clear what/when distinction without padding.

DimensionReasoningScore

Specificity

The description lists several concrete behaviors — 'Run a Cursor Composer review as an advisory second opinion', 'keeping the primary review verdict on the host brain', and 'Cursor never owns APPROVE/REQUEST_CHANGES' — matching the 'lists multiple specific concrete actions' anchor rather than the partial-action anchor 2.

3 / 3

Completeness

It explicitly answers what (advisory second opinion, host owns the verdict) and when via an explicit 'Use when user invokes cursor:review...' trigger clause, satisfying the both-what-and-when anchor 3 rather than anchor 2's implied-when.

3 / 3

Trigger Term Quality

It includes natural phrasings a user would actually say — 'user invokes cursor:review', 'asks Cursor to review', 'wants composer to sanity-check a diff' — giving good coverage, not just the single keyword of anchor 2.

3 / 3

Distinctiveness Conflict Risk

It carves a clear niche — an advisory Cursor lane where the host owns the verdict — with Cursor-specific triggers, making overlap with general review skills unlikely per anchor 3 rather than the still-overlapping anchor 2.

3 / 3

Total

12

/

12

Passed

Validation

93%

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

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

referenced_paths_exist

Referenced path issues: 2 missing

Warning

Total

15

/

16

Passed

Repository
Chachamaru127/claude-code-harness
Reviewed

Table of Contents

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.