CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

sce-commit

Analyze staged changes and run the regular or explicit bypass commit workflow

Quality

72%

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 ./.pi/skills/sce-commit/SKILL.md

The canonical home for this skill is sce-commit in crocoder-dev/shared-context-engineering

SKILL.md
Quality
Evals
Security

Quality

Content

81%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 is a well-engineered control-flow document: unambiguous sequencing, exact commands, explicit gates and stop conditions, and a clean separation between the workflow and the phase it dispatches to. Its main weaknesses are redundant guardrail repetition across sections and a second-level reference (commit-message-style.md) that is never surfaced from SKILL.md itself.

Suggestions

State "Read `references/atomic-commit.md` before running the phase" once (in the Phase reference section) and drop the three other repetitions in the workflow steps and Rules — a single authoritative statement earns the same compliance for fewer tokens.

Trim the Rules section to only the rules not already stated inline at the point of use (e.g. the amend/reset/push and unstaged-input prohibitions), rather than restating path-level guardrails a second time.

Surface `references/commit-message-style.md` from SKILL.md (one line in the Phase reference noting where message style lives) so every bundle file is discoverable from the top level.

DimensionReasoningScore

Conciseness

The body is directive throughout with no concept-explanation padding, but the same guardrails are repeated: "Read `references/atomic-commit.md` before running the phase" appears in the Phase reference, in both workflow paths, and again in Rules, and the Rules section restates prohibitions already stated inline ("Do not stage, unstage, restore, or otherwise modify repository or worktree files"). This is anchor 3 (mostly efficient but could be tightened) rather than 4, where only minor trims would be needed.

3 / 5

Actionability

The guidance is fully executable for an instruction-only skill: exact commands with exit-status semantics ("Run `git diff --cached --quiet`. A zero exit status means nothing is staged"), precise input parsing ("exactly `oneshot` or `skip`, compared case-insensitively"), explicit branch conditions on `status` values, and named output layouts. There are no vague steps; each action specifies what to run, what to check, and where to stop.

5 / 5

Workflow Clarity

Both paths are clearly sequenced with explicit validation checkpoints: the regular path gates on a user-confirmed staging prompt ("stop and prompt the user with the Regular-mode staging gate"), the bypass path validates staged content before proceeding, `status` branches are enumerated with defined stops, and failure handling is explicit ("Do not proceed past a failed `git commit`", "On failure, render the Bypass Git failure layout ... and stop"). This matches anchor 5's clear sequence with explicit validation and defined error handling.

5 / 5

Progressive Disclosure

The split is well designed — SKILL.md holds control flow ("This document holds the control flow ... the reference holds the phase itself") with phase details in `references/atomic-commit.md` and all layouts in `references/output.md`, both real files, clearly signaled, and read-order guidance is given. However `references/commit-message-style.md` is only reachable through `atomic-commit.md`, making the reference chain two levels deep; that minor organization gap holds this at anchor 4 rather than 5 ("one-level-deep references").

4 / 5

Total

17

/

20

Passed

Description

50%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 is concise and names a real domain, but it reads more like an internal workflow label than a skill description: it has a clear "what", no "when", and omits the concrete outputs (atomic commit messages, split proposals) and mode tokens that would make it discoverable and distinct. It scores a uniform 3 across all dimensions.

Suggestions

Add an explicit trigger clause, e.g. "Use when the user asks to commit staged changes, write commit messages, or says 'oneshot'/'skip'" — this would lift completeness and trigger_term_quality together.

State the concrete outputs in the "what": proposing atomic commit messages (regular mode) or producing and executing exactly one commit (bypass mode), instead of the abstract "run the ... workflow".

Include natural synonyms such as "commit message", "atomic commits", and "git commit" to distinguish this skill from generic git helpers.

DimensionReasoningScore

Specificity

"Analyze staged changes and run the regular or explicit bypass commit workflow" names the domain (staged changes, commits) and two actions (analyze, run workflow), but does not list what the workflow produces (commit messages, split proposals). This matches anchor 3 (domain plus 1-2 concrete actions); it falls short of anchor 4, which requires several specific actions.

3 / 5

Completeness

The "what" is clear (analyze staged changes, run a commit workflow in regular or bypass mode), but there is no "when" or equivalent trigger guidance anywhere in the description. Per the judging guidelines, a missing "Use when..." clause caps completeness at 3.

3 / 5

Trigger Term Quality

"staged changes", "commit", and "bypass" are natural user terms, but common trigger phrases such as "commit message", "atomic commit", or the mode tokens themselves ("oneshot", "skip") are absent. This sits at anchor 3 (some relevant keywords, missing common variations) rather than 4.

3 / 5

Distinctiveness Conflict Risk

"regular or explicit bypass commit workflow" gives it a niche, but "Analyze staged changes" overlaps heavily with generic commit-message and git-helper skills. Anchor 3 (somewhat specific, could still overlap with similar skills) fits better than 4 because the distinguishing modes are only weakly surfaced.

3 / 5

Total

12

/

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

crocoder-dev/shared-context-engineering

Path

.pi/skills/sce-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.