← 返回 Skills 市场

autonomous-loops

Name: autonomous-loops
Author: djc00p

作者 Deonte Cooper · GitHub ↗ · v1.0.1 · MIT-0

linuxdarwinwin32 ✓ 安全检测通过

127

总下载

当前安装

版本数

在 OpenClaw 中安装

/install autonomous-loops

功能描述

Autonomous Claude Code loop patterns: sequential pipelines, persistent REPL sessions, parallel spec-driven generation, PR automation, cleanup passes, and RFC...

使用说明 (SKILL.md)

Autonomous Loops — Patterns for Claude Code Automation

Running Claude in loops enables spec-driven development, CI/CD-style pipelines, and iterative refinement without human intervention between steps.

Quick Start

Choose your pattern by complexity:

Sequential Pipeline (simple) — Chain claude -p calls for linear workflows
Persistent REPL (simple) — Interactive sessions with history
Spec-Driven Parallel (medium) — Deploy N agents from spec, manage waves
PR Automation Loop (medium) — PR creation, CI fix, auto-merge
De-Sloppify Pass (add-on) — Cleanup step after any implementation
RFC-Driven DAG (advanced) — Multi-unit parallel work with dependency graph

Pattern Spectrum

Pattern	Setup	Complexity	Best For
Sequential Pipeline	Bash script	Low	Daily tasks, scripted workflows
REPL	Node/CLI	Low	Interactive development
Parallel Agents	Claude Code loop	Medium	Content generation, spec variations
PR Loop	Shell script	Medium	Iterative multi-day projects
De-Sloppify	Add-on to any	Optional	Quality cleanup after implementation
DAG Orchestration	Python/Node	High	Large features, parallel units, merge coordination

References

references/sequential-pipeline.md — Basic claude -p loops with examples
references/persistent-repl.md — NanoClaw-style session persistence
references/parallel-agents.md — Spec-driven deployment with wave management
references/pr-automation.md — Continuous Claude PR loop with CI gates
references/de-sloppify.md — Quality cleanup pattern
references/dag-orchestration.md — RFC-driven multi-unit coordination

Key Principles

Isolation — Each loop iteration gets fresh context (no bleed-through)
Context Persistence — Use files (SHARED_TASK_NOTES.md) to bridge iterations
Exit Conditions — Always set max-runs, max-cost, max-duration, or completion signal
No Blind Retries — Capture error context for next iteration
Separate Concerns — Different loop patterns for different problem sizes

Decision Matrix

Is this a single focused change?
├─ Yes → Sequential Pipeline
└─ No → Do you have a spec/RFC?
         ├─ Yes → Do you need parallel work?
         │        ├─ Yes → DAG Orchestration
         │        └─ No → PR Automation Loop
         └─ No → Do you need many variations?
                  ├─ Yes → Parallel Agents + Spec
                  └─ No → Sequential Pipeline + De-Sloppify

Anti-Patterns

❌ Infinite loops without exit conditions ❌ No context bridge between iterations ❌ Retrying the same failure without capturing error context ❌ Negative instructions instead of cleanup passes ❌ All agents in one context window (reviewer should never be the author) ❌ Ignoring file overlap in parallel work

Adapted from everything-claude-code by @affaan-m (MIT)

安全使用建议

This skill is coherent with its stated goal (automating Claude-driven pipelines), but it enables high-impact actions: creating branches, committing, pushing, opening and merging PRs, and auto-fixing CI failures. Before using it: 1) test in a sandbox repository and enable --disable-commits or dry-run modes; 2) set conservative caps (max-runs, max-cost, max-duration) and require manual approval before merges; 3) ensure your gh CLI is scoped to only the repos you intend (use a token with minimal scopes or a machine/service account); 4) review and control any session files (~/.claude/sessions) for sensitive data before allowing automation to read them; 5) prefer human review gates for critical repos; and 6) audit any scripts you run that they don't pipe remote content directly to bash. If you want tighter safety, require human-in-the-loop checks before merge or disable autonomous invocation for this skill.

功能分析

Type: OpenClaw Skill Name: autonomous-loops Version: 1.0.1 The skill bundle provides a comprehensive set of architectural patterns and documentation for automating Claude Code workflows, including sequential pipelines, parallel agent execution, and PR automation. The content consists of instructional markdown files and bash script examples that align with the stated purpose of enhancing autonomous development loops. There is no evidence of malicious intent, data exfiltration, or unauthorized execution; in fact, the documentation includes security-conscious advice such as warning against piping scripts directly to bash in references/pr-automation.md.

能力标签

cryptorequires-oauth-token

能力评估

✓ Purpose & Capability

The name/description (autonomous loops, pipelines, PR automation, DAG orchestration) aligns with the declared requirements: gh/git for repository and PR operations, node for the persistent REPL script examples, and CLAW_SESSION/CLAW_SKILLS to select session context/skill sets. Nothing requested appears unrelated to the described automation patterns.

ℹ Instruction Scope

SKILL.md instructs the agent to read/write session files (~/.claude/sessions/{name}.md), create/update SHARED_TASK_NOTES.md, run claude -p, run git/gh commands (create branches, push, create and merge PRs), and poll CI logs. These actions are within the skill's scope but are powerful: they will read local repo files, modify code, commit/push, and can auto-fix and auto-merge PRs if enabled. The doc correctly advises limits, but users should be aware these instructions grant the agent effective control over repo changes when invoked.

✓ Install Mechanism

No install spec and no code files are present; this is instruction-only. That lowers supply-chain risk because nothing is downloaded or written by an installer. The skill relies on existing binaries on PATH (gh, git, node).

✓ Credentials

Only CLAW_SESSION and CLAW_SKILLS are required (plus gh/git/node binaries). Those env vars map to session selection and skill composition and make sense for persistent REPL and pattern selection. The skill does not request unrelated credentials (no cloud keys, tokens, or passwords). Note: using gh implies the user will have gh auth configured — that grants repository privileges, so treat gh credentials carefully.

✓ Persistence & Privilege

always is false and the skill is user-invocable; it does not request permanent forced inclusion. The instructions describe persisting session and task files and automating PR lifecycle, which is expected behavior for this class of skill. Because autonomous invocation is allowed, the usual caution applies: if the agent is allowed to run these loops unattended, it may create/merge PRs and spend tokens/costs unless caps are set.

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install autonomous-loops
安装完成后，直接呼叫该 Skill 的名称或使用 /autonomous-loops 触发
根据 Skill 的参数说明提供必要输入，即可获得结构化输出

版本历史

v1.0.1

Fixed metadata: added required bins (gh, git, node) and env vars (CLAW_SESSION, CLAW_SKILLS) to requires

v1.0.0

Initial release. Autonomous agent loop patterns with quality gates and context management. Adapted from everything-claude-code by @affaan-m (MIT)

元数据

Slug autonomous-loops

版本 1.0.1

许可证 MIT-0

累计安装 1

当前安装数 1

历史版本数 2

常见问题