CtrlK
BlogDocsLog inGet started
Tessl Logo

diff-review

Render the patch-edit run's accumulated changes as a reviewable diff, surface it through a GenUI choice surface, and persist the user's accept / reject decision into the artifact manifest.

52

Quality

58%

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

Fix and improve this skill with Tessl

tessl review fix ./plugins/_official/atoms/diff-review/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

65%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.

A compact, well-organized spec that clearly defines inputs, outputs, and completion semantics. Its weaknesses are the missing executable steps for producing the diff artifacts and the absence of verification around the destructive rollback.

Suggestions

Add the concrete command(s) for producing the artifacts, e.g. `git diff > review/diff.patch`, so the generation step is executable rather than implied.

Add an explicit post-rollback verification step (e.g., confirm the working tree matches the pre-patch state or that only files in plan/steps.json files[] were touched) to close the loop on the destructive `git restore`.

Trim the rhetorical opener ('a code-migration / tune-collab handoff is worthless without a human-reviewable diff') to a one-line purpose statement.

DimensionReasoningScore

Conciseness

The body is a lean spec with no explanation of concepts Claude already knows; only the rhetorical opener ('a code-migration / tune-collab handoff is worthless without a human-reviewable diff') and the 'present the patch, capture the decision' gloss are trimmable. Anchor 4 ('efficient; minor instances... that could be trimmed'), not 5, because those two flourishes don't earn their tokens.

4 / 5

Actionability

Concrete artifacts and schemas are specified (the review/ output tree, decision.json's fields, `git restore`) and implementation paths are given, but the actual executable steps — how to generate diff.patch and summary.md — are never stated. Anchor 3 ('some concrete guidance but incomplete; missing key details'), not 4, because a core step has no command or code.

3 / 5

Workflow Clarity

The sequence is clear across sections (render diff → raise surface → capture decision → write manifest or rollback) with an explicit completion criterion, but the destructive `git restore` rollback has a stated boundary contract and no verification loop after it — the judging guidelines cap workflow clarity at 3 for destructive operations without validation. Not 4 because no post-rollback check exists.

3 / 5

Progressive Disclosure

The body is under 50 lines, no bundle files exist to reference, and it is organized into clean single-purpose sections (Inputs, Output, Convergence, Anti-patterns, Status). Per the simple-skill exception, well-organized sections alone merit 5; no nested or buried references exist.

5 / 5

Total

15

/

20

Passed

Description

51%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 states concrete, multi-step actions with a clear 'what', but it is written entirely in system-internal jargon with no user-facing trigger guidance. It would rarely fire from a natural user request.

Suggestions

Add a 'Use when...' clause naming the situations that should trigger this skill (e.g., 'Use when a patch-edit run finishes and the changes need user sign-off before continuing').

Replace internal jargon ('GenUI choice surface', 'artifact manifest', 'patch-edit run') with user-natural terms like 'show the diff', 'ask the user to accept or reject the changes', and mention concrete artifacts (diff.patch, decision.json).

Mention the per-file partial-accept flow so the capability coverage matches what the skill actually does.

DimensionReasoningScore

Specificity

Names three concrete actions — 'Render... as a reviewable diff', 'surface it through a GenUI choice surface', 'persist the user's accept / reject decision into the artifact manifest' — but coverage has gaps: the per-file partial flow and rollback, which the body treats as core, are absent. Anchor 4 ('several specific actions; minor gaps') fits better than 5.

4 / 5

Completeness

The 'what' is clear and concrete, but there is no 'Use when...' clause or equivalent trigger guidance, which caps completeness at 3 per the judging guidelines. Not 4 because the 'when' is entirely absent rather than merely implicit-but-stated.

3 / 5

Trigger Term Quality

Nearly all terms are internal jargon ('GenUI choice surface', 'artifact manifest', 'patch-edit run'); only 'diff' and 'accept / reject decision' read as words a user might naturally say. Anchor 2 ('one or two generic keywords; missing the natural phrases users say'), not 1, because 'diff' is genuinely user-natural.

2 / 5

Distinctiveness Conflict Risk

It carves a distinct niche (rendering a run's accumulated patch and capturing an accept/reject decision) with little overlap risk against generic skills, but 'diff', 'review', and 'accept/reject' language overlaps code-review skills. Anchor 4 ('mostly distinct; minor overlap risk'), not 5, since the triggers are not distinctive enough to prevent that overlap.

4 / 5

Total

13

/

20

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.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository
nexu-io/open-design
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.