CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

review-and-commit

Use when the user asks to review the uncommitted changes and then commit them — the verdict gates the commit. Not for a review alone or a plain commit. Self-contained — execute the inlined steps directly, do NOT invoke other skills.

58

Quality

73%

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/flowai/skills/review-and-commit/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

62%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.

An operationally rigorous composite with excellent sequencing, gate logic, validation, and error recovery — the workflow engineering is near best-in-class. It is heavily penalized by severe rhetorical padding and a monolithic all-inline structure with no progressive disclosure, and undermined by repeated citations of undefined Rules 10–13.

Suggestions

Cut the repeated restatements: the delegation-availability rule, the Phase 2 git-diff ban, and the scratch-cleanup warning each appear 2-3 times — state each rule once in the Rules section and reference it by number.

Define or inline Rules 7–13 (currently cited as "Rule 10", "Rule 11", "Rule 12", "Rule 13" throughout but never defined), and define the `<SCRATCH>`/`<sid>` placeholders the commands depend on.

Move the JiT methodology (steps 2b, 3d-3e, 8a-8b) and the Phase 3 reflection workflow into reference files loaded only when their gates fire, cutting the always-loaded body substantially while keeping the composite self-contained at the top level.

DimensionReasoningScore

Conciseness

The body is noticeably verbose with several padded sections: the Parallel Delegation bullet restates the "listing before inline check" and "speed is not one of the two conditions" rules three times each, Phase 2's git-diff ban is explained in three separate restatements, and motivational asides ("The claim is a factual report about this machine, and without the command's output behind it you are stating something you did not check") lecture rather than instruct. It avoids anchor 1 only because the material is genuinely non-obvious domain procedure rather than concepts Claude already knows.

2 / 5

Actionability

Guidance is overwhelmingly concrete — exact commands (`git worktree add <SCRATCH>/jit-parent-<sid> <parent-sha>`, `git log -1 --pretty=%B`, `git reset --hard HEAD~1`), an ordered first-match-wins verdict selection rule, and a full report template. Minor gaps keep it at anchor 4 rather than 5: the body repeatedly cites "Rule 10", "Rule 11", "Rule 12", and "Rule 13" which the Rules section never defines (it stops at 6), and `<SCRATCH>`/`<sid>` are undefined placeholders.

4 / 5

Workflow Clarity

Three strictly sequenced phases with an explicit verdict gate between them, error-recovery feedback loops (degradation notes, fallback rules, the Phase 3 check-two-things-then-reset undo procedure), and a closing verification checklist. Destructive operations (hard reset, commits, staging) all carry explicit validation checkpoints — exactly anchor 5.

5 / 5

Progressive Disclosure

No bundle files exist at all; three entire workflows (review, commit, reflect — the generator header confirms they are three separate atom source files) are inlined into one ~350-line monolith. Internal structure is real (headers, numbered steps, tagged blocks), which lifts it above anchor 2's "minimal structure", but content that clearly belongs in separate files is inline with no navigation, matching anchor 3.

3 / 5

Total

14

/

20

Passed

Description

73%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 with an explicit trigger clause, concrete negative boundaries that sharply distinguish it from review-only or commit-only skills, and third-person imperative voice. The main weakness is that the "what" side summarizes only the top-level two-phase action without hinting at the composite's actual capabilities.

Suggestions

Add one clause enumerating the core capabilities (e.g., "runs QA and code review, syncs docs, groups changes, and commits with Conventional Commits") so the what-side matches the strength of the when-side.

DimensionReasoningScore

Specificity

"review the uncommitted changes and then commit them" names the domain and two concrete actions gated by a verdict, but does not enumerate the composite's sub-capabilities (QA, code review, doc sync, reflection). This matches the anchor for 1-2 concrete actions without comprehensive coverage, and falls short of anchor 4 which expects several specific actions listed.

3 / 5

Completeness

Both what ("review the uncommitted changes and then commit them — the verdict gates the commit") and when (explicit "Use when..." clause with negative boundary) are present. The when is excellent, but the what stays thin on what the phases actually do, matching anchor 4 (both present, one could be more explicit) rather than anchor 5's fully concrete both.

4 / 5

Trigger Term Quality

Natural user phrases are present — "review the uncommitted changes", "commit", plus exclusion triggers "Not for a review alone or a plain commit". A few natural variations users might say ("check my diff", "land/ship my changes") are missing, matching anchor 4 rather than anchor 5's comprehensive synonym coverage.

4 / 5

Distinctiveness Conflict Risk

"Not for a review alone or a plain commit" explicitly disambiguates this composite from both single-purpose siblings, giving it a clear niche with distinct triggers and minimal conflict risk — exactly anchor 5.

5 / 5

Total

16

/

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

referenced_paths_exist

Referenced path issues: 3 missing

Warning

Total

15

/

16

Passed

Repository

korchasa/flowai-plugins

Path

plugins/flowai/skills/review-and-commit/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.