CtrlK
BlogDocsLog inGet started
Tessl Logo

worktree-hygiene

Safely reap merged Git worktrees and branches across Epicenter harnesses. Use when cleaning `git worktree list`, removing orphaned or detached worktrees, or deciding whether one is safe to delete, not when creating branches or commits.

72

Quality

91%

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

92%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 excellent single-file operational skill: dense, executable, and safety-gated, with a real decision procedure rather than vague advice. The only defect is two stray '</content></invoke>' wrapper tags left at the end of the file.

Suggestions

Remove the stray '</content>' and '</invoke>' lines (75-76) at the end of SKILL.md — they are generation artifacts that waste tokens and could confuse parsing.

The 'Related Skills' links point outside the bundle (../git/SKILL.md, ../standalone-commits/SKILL.md, ../pull-request/SKILL.md); either verify these paths exist in the deployment context or drop the ones that don't.

The audit loop's 'git cherry ... | grep -c ^+' output is used only for the unmerged label; a one-line note on interpreting 'ahead' counts for stacked branches would tighten the reap decision further.

DimensionReasoningScore

Conciseness

The body is lean and assumes Claude's git competence throughout — no space is spent explaining what worktrees or merge-base are ('Four harnesses (warp, codex, opencode, conductor) mint worktrees here and nothing reaps them', 'lockfile-only churn is throwaway'). Not a 5 only because the file ends with two stray wrapper tags ('</content>', '</invoke>' at lines 75-76) that are pure noise and could be trimmed.

4 / 5

Actionability

Fully executable guidance: a copy-paste audit pipeline ('git worktree list --porcelain | awk ... | while read b; do ... git cherry "$base" "$b" ... git merge-base --is-ancestor "$b" "$base"'), and a decision table with exact commands ('git worktree remove <path>; git branch -d/-D <branch>', 'git -C <path> switch -c holding/<name>'). Not below 5: placeholders are appropriate parameterization, not pseudocode.

5 / 5

Workflow Clarity

Destructive/batch skill with explicit validation checkpoints and feedback loops: fetch-first, self-run audit ('Never reap on a self-reported "it's merged"; run the audit yourself'), an approval gate before any destructive command, and a Final Checks section ('Re-run the audit; confirm reaped paths are gone', 'git worktree prune', 'Confirm any anchored holding/* branches still hold their edits'). This clears the destructive-operations cap comfortably.

5 / 5

Progressive Disclosure

A single-file skill (no references/, scripts/, or assets/ directories exist, and none are needed at this size) with well-organized, navigable sections (reap signal, audit procedure, decision table, gotchas, invariant, final checks) and one clearly-signaled Related Skills pointer. Content is appropriately placed; nothing is inlined that belongs in a separate file.

5 / 5

Total

19

/

20

Passed

Description

87%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: concrete actions, explicit 'Use when...' triggers with a negative boundary, and a distinct niche. Its only weakness is vocabulary — 'reap' is coined jargon and common cleanup synonyms are missing, which slightly lowers natural-trigger matching.

Suggestions

Add natural synonyms users would say, e.g. 'Use when cleaning up, deleting, or pruning old worktrees' so the trigger vocabulary matches what users actually type.

Consider glossing the coinage on first use ('reap (remove) merged worktrees and their branches') so the core action verb is immediately legible to both the model and the user.

DimensionReasoningScore

Specificity

Several concrete actions are named — 'reap merged Git worktrees and branches', 'cleaning git worktree list', 'removing orphaned or detached worktrees', 'deciding whether one is safe to delete'. Not 5: coverage is strong but misses some adjacent concrete operations (e.g., pruning stale registrations) and 'reap' leans on context for meaning; not 3: more than 1-2 specific actions are listed with clear scope.

4 / 5

Completeness

Explicitly answers both questions: what ('Safely reap merged Git worktrees and branches across Epicenter harnesses') and when ('Use when cleaning git worktree list, removing orphaned or detached worktrees, or deciding whether one is safe to delete'), with a concrete negative boundary ('not when creating branches or commits'). Matches the top anchor exactly.

5 / 5

Trigger Term Quality

Good natural-keyword coverage: 'git worktree list', 'orphaned', 'detached worktrees', 'safe to delete', plus branch/merge terms. Not 5: 'reap' is the skill's own coinage rather than a phrase users would naturally say, and common synonyms like 'clean up worktrees', 'delete worktrees', or 'prune' are absent; not 3: multiple natural trigger phrases users would actually utter are present.

4 / 5

Distinctiveness Conflict Risk

Clear niche (merged-worktree/branch cleanup) with distinct triggers and an explicit exclusion of the adjacent task ('not when creating branches or commits'), which cleanly separates it from general git/commit/PR skills. Minimal conflict risk.

5 / 5

Total

18

/

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

relative_links

Relative link issues: 3 suspicious

Warning

Total

15

/

16

Passed

Repository
EpicenterHQ/epicenter
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.