CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

gitnexus-review

Review code changes with GitNexus from a GitHub PR URL or number, a branch/ref or commit range, or local staged, unstaged, and untracked changes. Use when the user asks for a code review, merge-risk assessment, regression hunt, missing-test analysis, or a verdict on whether a PR, branch, commit range, or local diff is safe.

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

The canonical home for this skill is gitnexus-review in abhigyanpatwari/GitNexus

SKILL.md
Quality
Evals
Security

Quality

Content

66%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 body delivers an unusually concrete, well-sequenced review workflow with genuine validation and error-recovery loops, but it is a monolithic ~280-line document whose GitNexus-internal reference material is inlined, and it points to a ci-personas/ bundle directory that does not exist. Splitting reference material into files and fixing or removing the broken lane reference would address the weakest dimensions.

Suggestions

Move the GitNexus-internals invariants (SCHEMA_FINGERPRINT derivation, parse-cache-identity ownership boundaries, benchmark fingerprint baselines) into a one-level-deep reference file (e.g. references/gitnexus-invariants.md) and link to it from workflow step 8 instead of inlining ~20 lines of repo-specific detail.

Fix the broken bundle reference: either ship the ci-personas/ lane files the Swarm lanes section depends on ("Six dispatchable lane definitions ship with this skill in ci-personas/") or remove the section's claims about copying ci-personas/*.md, since the directory is absent from the bundle.

Tighten the winding multi-clause prose (e.g. the checkout/index alignment and expert-lens sections) into short imperative bullets to cut token cost while preserving the concrete guidance.

DimensionReasoningScore

Conciseness

The body is dense and mostly earns its tokens, but the ~20-line GitNexus-internals digression ("SCHEMA_FINGERPRINT... NODE_SCHEMA_QUERIES + REL_SCHEMA_QUERIES... benchmark fingerprint sets... parse-cache-identity.ts") and repeated evidence-requirement caveats are more than minor over-explanation, matching anchor 3 rather than 4.

3 / 5

Actionability

Concrete commands and parameterized tool invocations throughout ("gh pr view", "git merge-base <base> <head>", "node .gitnexus/run.cjs analyze --index-only", "detect_changes" with scope/base_ref) give mostly executable guidance with only minor gaps such as full gh --json invocations, matching anchor 4.

4 / 5

Workflow Clarity

The numbered 8-step workflow has explicit validation checkpoints ("If an exact target checkout/index cannot be established, state the limitation"), error-recovery loops ("On DEFECTS, repair the draft and re-dispatch the critic once"), and fallbacks ("run that lane's charge inline"), matching the anchor 5 pattern.

5 / 5

Progressive Disclosure

The ~280-line monolith inlines content that clearly belongs in separate files (lane definitions, GitNexus schema-fingerprint invariants), and its sole bundle reference is broken: the body claims "Six dispatchable lane definitions ship with this skill in ci-personas/" but that directory is absent from the actual bundle, matching anchor 2.

2 / 5

Total

14

/

20

Passed

Description

91%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.

A strong description: third-person, concrete about what it reviews and from which surfaces, with an explicit and well-phrased 'Use when' trigger clause covering the natural ways users ask for this. The only weaknesses are the single-verb 'what' and mild overlap risk on the broad 'code review' trigger.

DimensionReasoningScore

Specificity

Enumerates concrete input surfaces ("GitHub PR URL or number, a branch/ref or commit range, or local staged, unstaged, and untracked changes") and deliverables ("merge-risk assessment, regression hunt, missing-test analysis, or a verdict"), but the core 'what' is a single review verb rather than a list of multiple distinct capabilities, matching anchor 4 rather than 5.

4 / 5

Completeness

Explicitly answers 'what' ("Review code changes with GitNexus from a GitHub PR URL or number...") and 'when' with a concrete "Use when the user asks for..." clause listing concrete trigger phrases, exactly matching the anchor 5 pattern.

5 / 5

Trigger Term Quality

Covers the natural phrases users would say for this task — "code review", "merge-risk assessment", "regression hunt", "missing-test analysis", "PR", "branch", "commit range", "local diff", "safe" — including synonyms and surface variants, matching the comprehensive anchor 5.

5 / 5

Distinctiveness Conflict Risk

The GitNexus-specific framing and enumerated diff surfaces give it a clear niche, but generic triggers like "code review" and "PR" overlap closely related review skills, matching anchor 4 rather than the minimal-conflict anchor 5.

4 / 5

Total

18

/

20

Passed

Validation

100%

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

abhigyanpatwari/GitNexus

Path

gitnexus/skills/gitnexus-review/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.