claude-code

Operate Claude Code CLI (v2.1+) via the terminal only when the user explicitly requests Claude Code or names this skill. Covers print mode (-p), interactive tmux sessions, and background (--bg) orchestration.

Quality

93%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

Quality

Content

100%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 exemplary instruction-only skill body: dense tables and executable commands, a clear setup-to-cleanup workflow with explicit validation signals, and a well-organized one-level-deep reference structure. No notable defects found on any content dimension.

Dimension	Reasoning	Score
Conciseness	The body is lean and dense throughout — a mode-selection table, terse bullet guardrails, and minimal one-line context ("Delegate coding work to Claude Code, Anthropic's autonomous coding agent CLI") with zero padding or explanation of concepts Claude already knows. Every token earns its place.	5 / 5
Actionability	Fully executable, copy-paste-ready commands cover the common cases: the primary `claude -p ... --allowedTools ... --max-turns ... --output-format json` invocation, five concrete variants (pipe input, resume, structured extraction, CI, cost cap), and the background-mode command set (`claude --bg`, `claude agents --json --all`, attach/logs/stop).	5 / 5
Workflow Clarity	The sequence is explicit and checkpointed: setup with verification (`claude auth status` / `claude doctor`), platform check before running, a "decide FIRST" mode-selection table, documented outcome signals (`subtype`: `success` \| `error_max_turns` \| `error_budget`), post-run verification ("summarize what changed (use the JSON `result` + `git diff`)"), and an 8-item guardrail checklist including "Don't kill slow sessions — capture progress first" before destructive cleanup.	5 / 5
Progressive Disclosure	The main file is a concise overview with a References table giving a "read when" condition for each of the five bundle files (all verified to exist: print-mode.md, interactive-tmux.md, flags.md, config.md, platforms.md), plus inline one-level-deep signals like "Full auth/install variants → references/flags.md". Navigation is easy and nothing that belongs in a reference is inlined.	5 / 5
	Total	20 / 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 that clearly and explicitly states both capability coverage (three concrete modes with flags) and a precise activation trigger in third person. The only weakness is that it lists coverage areas rather than concrete task-level actions, leaving minor specificity and keyword gaps.

Suggestions

Add one or two task-level actions (e.g., "run one-shot fixes, diff reviews, or structured extraction") to raise specificity from mode enumeration to concrete capability actions.

Include common trigger synonyms such as "claude CLI" or "delegate coding to Claude" so users phrasing the request differently still match naturally.

Dimension	Reasoning	Score
Specificity	Names the domain and enumerates specific concrete coverage ("print mode (-p), interactive tmux sessions, and background (--bg) orchestration"), but task-level actions such as fixing bugs, reviewing diffs, or delegating coding work are not listed, leaving minor gaps versus comprehensive coverage.	4 / 5
Completeness	Explicitly answers both what ("Operate Claude Code CLI (v2.1+) via the terminal" with the three modes) and when ("only when the user explicitly requests Claude Code or names this skill") with a concrete trigger phrase.	5 / 5
Trigger Term Quality	Includes natural trigger terms users would say ("Claude Code", "print mode (-p)", "tmux", "background (--bg)") with an explicit request condition, but misses common synonyms like "claude CLI", "delegate to claude", or "terminal agent".	4 / 5
Distinctiveness Conflict Risk	Clear niche (driving the Claude Code CLI specifically) with an explicit, narrow trigger that requires the user to name Claude Code or the skill, minimizing overlap with generic coding-assistant skills.	5 / 5
	Total	18 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: XiaomiMiMo/MiMo-Code
Path: packages/opencode/src/skill/builtin/.bundle/claude-code/SKILL.md
Commit: f0293bf

Reviewed: 2 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.