autonomous-loops

自主运行 Claude Code 循环的模式与架构 —— 从简单的顺序流水线到 RFC 驱动的多智能体 DAG 系统。

Quality

67%

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

Fix and improve this skill with Tessl

tessl review fix ./skills/autonomous-loops/SKILL.md

Quality

Content

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable with strong, well-validated workflows and rich executable examples. Its weaknesses are conciseness (motivational/rationale prose and length) and progressive disclosure (a 600-line monolith with no reference files despite being a complex multi-pattern skill).

Suggestions

Split each loop pattern into its own reference file (e.g. references/sequential-pipeline.md, references/ralphinho-dag.md) and keep SKILL.md as a concise overview with one-level-deep links.

Trim motivational rationale (the '核心见解' callouts and the '为什么不能使用负面指令' essay) to the essential rule, removing prose that restates what Claude already knows.

Move the long flag/reference tables (e.g. Continuous Claude config, complexity tiers) into their respective reference files to reduce inline token load.

Dimension	Reasoning	Score
Conciseness	The body is dense with concrete, executable code and diagrams, but it also carries motivational/rationale prose (e.g. '如果你无法理解这样的循环...', the '为什么不能使用负面指令' essay) and re-explains behaviors Claude already knows, so it could be tightened rather than earning the lean level-3 anchor.	2 / 3
Actionability	It provides fully executable bash scripts, copy-paste `claude -p`/`continuous-claude` invocations, a ready-to-use `.claude/commands/infinite.md` template, and a TypeScript WorkUnit interface — concrete and copy-paste ready as the level-3 anchor requires.	3 / 3
Workflow Clarity	Multi-step processes are explicitly sequenced with validation checkpoints and feedback loops: `set -e` exit-code propagation, CI-failure recovery (gh run list → fix → retry with --ci-retry-max), and the merge-queue eviction → re-enter-loop-with-context cycle, plus an anti-pattern checklist.	3 / 3
Progressive Disclosure	There are no bundle files (references/scripts/assets) and the ~600-line document is a single monolithic file with good internal navigation (ToC, anchor links, decision matrix), but content that could be split into per-pattern reference files is all inline, matching the level-2 'content that should be separate is inline' anchor.	2 / 3
	Total	10 / 12 Passed

Description

57%

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 identifies a clear, distinctive niche and names specific architectures, but it is noun-heavy (patterns rather than actions), lacks natural trigger variations, and omits an explicit 'Use when' clause. It sits solidly at level 2 on most dimensions.

Suggestions

Add an explicit 'Use when...' trigger clause, e.g. 'Use when setting up autonomous/unattended Claude Code workflows, CI-style dev pipelines, or parallel multi-agent orchestration.'

Lead with concrete verbs/actions (e.g. 'Build, sequence, and orchestrate autonomous Claude Code loops') rather than the abstract 'patterns and architectures'.

Broaden trigger terms with common phrasings users would say ('run Claude unattended', 'self-driving dev loop', 'CI for Claude Code').

Dimension	Reasoning	Score
Specificity	The description names the domain ('自主运行 Claude Code 循环') and two specific patterns ('顺序流水线', 'RFC 驱动的多智能体 DAG'), but frames them as 'patterns and architectures' rather than concrete actions, so it is not a comprehensive action list as the level-3 anchor requires.	2 / 3
Completeness	It clearly answers 'what' (patterns/architectures for autonomous loops) but has no 'Use when...' clause or equivalent explicit trigger guidance, which the guidelines cap at 2.	2 / 3
Trigger Term Quality	It includes some natural terms a user might say ('Claude Code 循环', '自主运行', '流水线'), but leans technical ('DAG', 'RFC 驱动') and omits common variations or an explicit 'Use when' trigger phrase.	2 / 3
Distinctiveness Conflict Risk	'Autonomous Claude Code loop orchestration' with concrete sub-patterns (sequential pipeline, RFC-driven multi-agent DAG) is a clear, narrow niche unlikely to conflict with unrelated skills.	3 / 3
	Total	9 / 12 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (608 lines); consider splitting into references/ and linking	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
referenced_paths_exist	Referenced path issues: 2 missing	Warning

	Total	13 / 16 Passed

Repository: xu-xiang/everything-claude-code-zh
Path: skills/autonomous-loops/SKILL.md
Commit: dfbf946

Reviewed: about 1 hour 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.