cursor-do

Delegate a single write task to Cursor Composer via cursor-companion.sh inside an isolated worktree, then Lead-review the diff and cherry-pick. Use when user invokes cursor:do, says delegate to cursor, have composer write it, refactor with cursor, hand a file edit to Composer. Do NOT load for: planning, code review only, read-only investigation, or multi-task team runs (use breezing --cursor or cursor:ask instead).

Quality

90%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

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.

A highly actionable, well-sequenced delegation workflow with strong validation gates and feedback loops for a destructive git operation. The main weakness is conciseness: the plugin-root resolution logic is duplicated across Step 3 and Step 5, and some narration-rule examples could be trimmed.

Suggestions

Extract the duplicated valid_root plugin-root resolution block into a shared script (e.g. scripts/resolve-plugin-root.sh) and call it from both Step 3 and Step 5 to remove ~25 lines of repetition.

Compress the Step 0 narration-rules section: keep one violation example and one normal example instead of the multi-block treatment, since the banner/plan contract is already concrete.

Consider moving the full cursor-companion prompt template (Step 5) into references/ and inlining only a short invocation, reducing the inline body length while keeping the workflow actionable.

Dimension	Reasoning	Score
Conciseness	The body is largely dense, actionable bash with little conceptual padding, but the plugin-root valid_root resolution block (~25 lines) is duplicated verbatim in Step 3 and Step 5, and the Step 0 narration-rules section repeats violation/normal examples, so it could be tightened.	3 / 5
Actionability	Every step ships complete, copy-paste-ready bash with concrete commands, guard conditions, and exit-code handling (e.g. the Step 2 parallel pre-check and Step 5 cursor-companion invocation), covering the common cases fully.	5 / 5
Workflow Clarity	Steps 0–8 are explicitly sequenced with validation checkpoints — Step 6 diff review + contract-grep gates, Step 7 precondition checks (commit count, clean Plans.md) — plus retry-once-then-abort feedback loops and a Full Containment checklist for this destructive cherry-pick-into-main operation.	5 / 5
Progressive Disclosure	Structure is clear (numbered steps, containment table, prohibited list) and the one bundle reference references/cursor-cli-only.md is well-signaled and one level deep; however the body itself is a fairly monolithic ~400-line inline workflow rather than an overview pointing into deeper materials.	4 / 5
	Total	17 / 20 Passed

Description

100%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 exemplary: concrete third-person actions, comprehensive natural trigger terms, explicit what/when coverage, and strong negative-scope disambiguation against sibling skills. No weaknesses to address.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions in third person — 'Delegate a single write task to Cursor Composer via cursor-companion.sh inside an isolated worktree, then Lead-review the diff and cherry-pick' — covering delegation, worktree isolation, diff review, and cherry-pick comprehensively.	5 / 5
Completeness	Explicitly answers both what (delegate write task, Lead-review, cherry-pick) and when ('Use when user invokes cursor:do, says delegate to cursor...'), plus negative-scope guidance, with concrete trigger phrases.	5 / 5
Trigger Term Quality	Covers natural phrasings a user would actually say — 'delegate to cursor', 'have composer write it', 'refactor with cursor', 'hand a file edit to Composer' — including synonyms and the explicit cursor:do invocation token.	5 / 5
Distinctiveness Conflict Risk	Clear niche (single-task write delegation with cherry-pick) and an explicit 'Do NOT load for' clause naming breezing --cursor and cursor:ask, minimizing overlap with adjacent skills.	5 / 5
	Total	20 / 20 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
relative_links	Relative link issues: 2 missing, 2 deeper-than-1-level	Warning
referenced_paths_exist	Referenced path issues: 4 missing, 1 deeper-than-1-level	Warning

	Total	14 / 16 Passed

Repository: Chachamaru127/claude-code-harness
Path: opencode/skills/cursor-do/SKILL.md
Commit: 2b2b748

Reviewed: 3 days ago

Table of Contents

Discovery Implementation Validation

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.