CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

review-change

Review a change.

Quality

36%

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 ./internal/preserve/testdata/markdown-generated/package/skills/review-change/SKILL.md

The canonical home for this skill is review-change in jbaruch/agentic-context-registry

SKILL.md
Quality
Evals
Security

Quality

Content

22%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.

The SKILL.md body is effectively empty — a single heading that restates the skill name. It contains no actionable guidance, no workflow, and nothing that would let Claude (or a user) understand what reviewing a change actually involves, making the skill unusable as written.

Suggestions

Add concrete instructions for the core task: how to obtain the change (e.g., git diff of the working tree or a PR number), what to look for, and how to report findings.

If the review targets code changes, include a short sequenced workflow with an explicit validation step (e.g., re-run checks on the diff, verify each finding against the actual changed lines before reporting).

Either fill the body with the actual review procedure or, if detail lives elsewhere, add the missing reference files and link them clearly from the body.

DimensionReasoningScore

Conciseness

The body ("# Review change") contains no padded explanations or over-explanation, but its only content — a heading that duplicates the name field — provides no instructional value; genuine token efficiency requires lean content that actually conveys guidance. It is not 4-5 because the tokens present do not earn their place, and not 1-2 because there is no verbosity to trim.

3 / 5

Actionability

The body provides no concrete code, commands, or steps — it does not even describe the task, let alone instruct how to perform it. This matches (and undershoots) the anchor 1 case of only describing rather than instructing.

1 / 5

Workflow Clarity

Steps are entirely missing — no sequence, no checkpoints, no validation — matching anchor 1 ("Steps missing or incoherent; no sequence"). The simple-skill exception does not apply because "Review change" is not an unambiguous single action: what kind of change and how to review it are undefined.

1 / 5

Progressive Disclosure

The body has no structural defects — no monolithic text, no buried or nested references, and no bundle files are misused — but it also has no organized sections or content to navigate. The simple-skill 5 exception requires well-organized sections, which a bare heading with zero body content is not; it is not a 2 because nothing inlined here belongs in a separate file.

3 / 5

Total

8

/

20

Passed

Description

20%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 "Review a change." is a two-word generic label rather than a capability description. It fails to state what the skill concretely does, when it should be used, or how it differs from the many other review-oriented skills a user might invoke.

Suggestions

State 2-3 concrete actions the skill performs (e.g., "Analyze a git diff for bugs, style issues, and unsafe patterns, then post inline review comments").

Add an explicit trigger clause, e.g., "Use when the user asks to review a change, diff, PR, or staged changes" — the missing 'Use when...' guidance currently caps completeness at 3.

Include distinguishing keywords for the change type in scope (code, git diff, pull request, working tree) to reduce conflict with other review skills.

DimensionReasoningScore

Specificity

The description is only "Review a change." — it names no domain and no concrete actions, which is entirely vague abstract language matching the anchor 1 example ("Helps with documents"). It is not a 2 because even that anchor expects a domain to be named ("Processes PDF files"), and no domain appears here.

1 / 5

Completeness

The description has a vague "what" ("Review a change.") and no "when" clause at all, matching anchor 2 exactly. It is not a 1 because it does state an action verb rather than pure filler, and not a 3 because anchor 3 requires a clear "what" — and the judging guidelines cap completeness at 3 when a "Use when..." clause or equivalent is missing.

2 / 5

Trigger Term Quality

"Review" and "change" are two generic keywords, missing the natural phrases users actually say ("code review", "review my diff", "PR review") — closest to the anchor 2 example ("Works with files"). It is not a 1 because the language is plain-generic rather than technical jargon, and not a 3 because no relevant keyword variations, synonyms, or file extensions appear.

2 / 5

Distinctiveness Conflict Risk

"Review a change." is very broad and would overlap with many similar skills — any code-review, security-review, or document-review skill conflicts with it. It is not a 1 because it does denote a task category (reviewing changes) rather than claiming to help with everything, and not a 3 because nothing in the text distinguishes it from those similar skills.

2 / 5

Total

7

/

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

description_field

'description' is very short (16 chars), consider making it more detailed

Warning

Total

15

/

16

Passed

Repository

jbaruch/agentic-context-registry

Path

internal/preserve/testdata/markdown-generated/package/skills/review-change/SKILL.md

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.