CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

git-checkpoint

Use git as a safety net - create a checkpoint commit before risky or large changes and roll back cleanly if a change makes things worse. Use before multi-file refactors.

Quality

88%

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

SKILL.md
Quality
Evals
Security

Quality

Content

87%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 exceptionally lean, fully executable skill body whose only real weakness is the absence of validation around its destructive rollback commands. Adding a verification step before and after `git reset --hard` would resolve the workflow-clarity cap.

Suggestions

Add a verification checkpoint before destructive rollback, e.g. "Before `git reset --hard`, run `git stash list` / `git status` and confirm nothing uncommitted is worth keeping" — this addresses the workflow_clarity cap for destructive operations.

Add a post-rollback validation step, e.g. confirm the tree is clean and at the checkpoint with `git status && git rev-parse HEAD` before retrying a different approach.

Include a brief note on verifying the checkpoint itself succeeded (e.g. check `git log -1` output) so a failed checkpoint commit cannot be mistaken for a good saved state.

DimensionReasoningScore

Conciseness

The body is lean and assumes Claude's competence: a two-line rationale, one executable checkpoint command, three one-line rollback commands, and two rules. Nothing explains git basics, and every token earns its place — a clean match for anchor 5; anchor 4 would require noticeable over-explanation to trim.

5 / 5

Actionability

All guidance is copy-paste executable: `git add -A && git commit -m "checkpoint: before <short description>" --no-verify`, `git rev-parse HEAD`, `git checkout -- <file>`, `git reset --hard HEAD`, and `git reset --hard <checkpoint-hash>` cover the common cases exactly, matching anchor 5. It is not anchor 4 because there are no gaps in the commands themselves.

5 / 5

Workflow Clarity

The sequence (checkpoint → record hash → roll back on failure → retry differently) is clear, but the workflow's core operations are destructive (`git reset --hard` discards all uncommitted changes) and there are no validation or verification steps — e.g., confirming the checkpoint commit succeeded, checking `git status`/stash for work worth keeping before a hard reset, or verifying the tree state after rollback. Per the rubric, missing validation in destructive-operation workflows caps this at 3, overriding the simple-skill exception.

3 / 5

Progressive Disclosure

This is a sub-50-line single-purpose skill with no external references needed, and the content is organized into clear headed sections (Checkpoint, Roll back, Rules) with nothing that belongs in a separate file. Per the rubric's simple-skill guideline, well-organized sections alone earn anchor 5.

5 / 5

Total

18

/

20

Passed

Description

83%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: concise, concrete, and follows the what-plus-when pattern with natural trigger phrases. The only improvements are broader trigger synonyms (undo, revert, snapshot) and slightly wider trigger coverage beyond multi-file refactors.

DimensionReasoningScore

Specificity

The description names the domain ("git as a safety net") and lists two concrete, conditioned actions — "create a checkpoint commit before risky or large changes" and "roll back cleanly if a change makes things worse" — which fully covers this skill's scope with only minor gaps (e.g., no mention of discarding uncommitted changes). It fits anchor 4 (several specific actions, minor gaps) better than anchor 3, whose example covers only 1-2 actions without conditions, and better than anchor 5, which expects a broader multi-action catalog.

4 / 5

Completeness

It explicitly answers both questions: what ("create a checkpoint commit... and roll back cleanly") and when ("Use before multi-file refactors", plus "before risky or large changes") with concrete trigger phrases. This matches anchor 5's pattern of a clear what followed by an explicit 'Use when...' clause; anchor 4's example has a vaguer, less conditioned 'when'.

5 / 5

Trigger Term Quality

Natural trigger phrases include "risky or large changes", "multi-file refactors", "roll back", and "checkpoint commit" — terms a user would plausibly say. It falls short of anchor 5 because common synonyms like "undo", "revert", or "snapshot" are missing, but it is well above anchor 3's level of missing common variations.

4 / 5

Distinctiveness Conflict Risk

The checkpoint/rollback safety-net niche is fairly distinct with its own triggers, but the broad use of "git" creates minor overlap risk with other git-related skills (commit messages, PR workflows). It is clearly above anchor 3 (overlap with similar skills) but not the fully minimal-conflict profile of anchor 5.

4 / 5

Total

17

/

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

agentscope-ai/agentscope-java

Path

agentscope-examples/agents/agentscope-codingagent/src/main/resources/workspace-templates/skills/git-checkpoint/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.