← 返回 Skills 市场
Auto Revolution
作者
Jaden's built a claw
· GitHub ↗
· v2.0.2
· MIT-0
127
总下载
0
收藏
1
当前安装
2
版本数
在 OpenClaw 中安装
/install auto-revolution
功能描述
Safe publishing variant of Auto Revolution. Provides structured task templates, review prompts, safety scanning, task state updates, and logging for human su...
使用说明 (SKILL.md)
Auto Revolution
This is the ClawHub publishing variant.
Use this skill for
- creating structured task JSON files
- activating queued tasks whose dependencies are complete
- scanning proposed instruction text for risky shell patterns
- applying review JSON back into a task file
- logging workflow events in JSONL
- generating a review prompt for manual or supervised review
Do not use this package for
- autonomous cron loops
- unsupervised background execution
- direct execution of arbitrary shell instruction strings
- remote agent spawning from bundled scripts
Files to use
task-schema.jsonfor task structure referenceconfig/models.jsonfor example role configurationscripts/create-task.jsto create a task interactivelyscripts/activate-queued-tasks.jsto move ready tasks fromqueuedtopendingscripts/security-scan.jsto scan instruction text or a task review blockscripts/trigger-review.jsto generate a review prompt for supervised reviewscripts/apply-review.jsto write a review result back into a task JSON filescripts/log-event.jsto append JSONL audit eventsscripts/atomic-lock.sh,scripts/force-unlock.sh,scripts/unblock-task.shfor task lock maintenance
Minimal workflow
- Create or edit a task JSON.
- If the task depends on other tasks, run
scripts/activate-queued-tasks.jswhen dependencies may be complete. - Generate a review prompt with
scripts/trigger-review.js. - Obtain a supervised review result as JSON.
- Apply that result with
scripts/apply-review.js. - If you need to inspect instruction text, run
scripts/security-scan.jsbefore any manual execution.
Safety rules
- Keep a human in the loop for any real execution.
- Treat generated instructions as untrusted until reviewed.
- Prefer structured task updates over free form shell execution.
- Do not add scripts that call
execSyncon model generated instruction strings in the publishing package.
Notes
This package is intentionally minimal for publication. Internal development documents, autonomous heartbeat runners, and direct execution engines belong in the private local version, not the published package.
安全使用建议
This package broadly matches a supervised task-review/publishing helper, but proceed cautiously:
- Do not assume the package is a drop-in complete system: several scripts referenced in README/SKILL.md are missing (start-reviewer.js, start-executor.js, iron-heartbeat.js) — the README and some outputs point to workflows that rely on external components not bundled here.
- Inspect and fix shell scripts before running: force-unlock.sh and unblock-task.sh appear to contain a stray trailing '}' and other minor syntax issues; test them in a safe environment first.
- Beware of implicit automation suggestions: trigger-review.js recommends using sessions_spawn (automatic invocation) as 'recommended'. If you want a strictly supervised flow, avoid using sessions_spawn or any automatic runner that would feed model output back without human review.
- The security scanner is helpful (security-scan.js), but it is pattern-based and not foolproof. Always review any generated 'next_instructions' yourself before executing them.
- Run the scripts in an isolated environment (sandbox/container) with backed-up task/log directories the first time, and validate behavior on sample tasks.
If you plan to use this in production or link it to hosted LLMs, ask the publisher for the missing runner scripts or documentation about the intended orchestration and ensure platform credentials are stored and used outside this package. Fix the reported script syntax issues and verify that automatic invocation paths are intentionally enabled and audited before enabling them.
功能分析
Type: OpenClaw Skill
Name: auto-revolution
Version: 2.0.2
The 'auto-revolution' skill bundle is a task management framework designed for human-supervised workflows. It provides structured utilities for creating tasks, managing state transitions (queued, pending, reviewing), and implementing file-based concurrency locking via 'scripts/atomic-lock.sh'. A key security feature is 'scripts/security-scan.js', which uses regex patterns to proactively detect dangerous shell commands like RCE payloads or unauthorized file deletions. The code is well-documented, lacks autonomous execution capabilities, and contains no evidence of data exfiltration or malicious intent.
能力评估
Purpose & Capability
The skill's name, description, and included scripts (task creation, queued activation, security scanning, review trigger, apply-review, logging, lock helpers) align with the advertised 'safe publishing' / supervised-review purpose. It does not request environment credentials or external config. Note: the package references LLM role names (e.g. advanced-model/sonnet, anthropic/claude-sonnet-4-6) but does not request or document API keys — this is plausible for a publishing bundle that expects a platform-level agent, but users should be aware it assumes an external agent/model infrastructure.
Instruction Scope
SKILL.md and scripts otherwise stay within the stated supervised workflow boundaries, but multiple instructions reference scripts or behaviors that are not present (e.g., create-task.js and README mention node scripts/start-reviewer.js and node scripts/start-executor.js; apply-review.js and create-task.js mention node scripts/iron-heartbeat.js). trigger-review.js explicitly recommends an 'automatic (recommended) sessions_spawn' flow to call Sonnet and then auto-apply results — this encourages autonomous invocation despite the package's repeated claim that it avoids autonomous background execution. These inconsistencies grant the agent/user broad discretion and could mislead someone expecting a purely manual, audited flow.
Install Mechanism
No install spec is provided (instruction-only publish bundle). All code is bundled in the package; nothing is downloaded or executed automatically during install, which is the lowest-risk install behavior.
Credentials
The skill requires no environment variables, credentials, or external config paths. It references model names only in config/models.json and prompts; these do not themselves require secrets in the bundle. This is proportionate to a local publishing helper package, but if you intend to integrate with hosted models you will need platform credentials that are not requested here.
Persistence & Privilege
The package does not request always:true or any persistent system-level privileges. It reads/writes files under local tasks/ and logs/ directories only. That level of persistence is expected for a task management helper and is consistent with the stated purpose.
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 auto-revolution - 安装完成后，直接呼叫该 Skill 的名称或使用
/auto-revolution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
v2.0.2
Publish safe minimal ClawHub variant with supervised workflow helpers only. Removed autonomous heartbeat runners, direct execution engine, review runner, internal design docs, and test scripts.
v2.0.1
Auto Revolution Skill 2.0.1 is a sanitized (脱敏) release.
- Public release with sensitive information removed.
- No functional or workflow changes since 2.0.0.
- Documentation improvements for clarity and safety.
元数据
常见问题
Auto Revolution 是什么？
Safe publishing variant of Auto Revolution. Provides structured task templates, review prompts, safety scanning, task state updates, and logging for human su... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 127 次。
如何安装 Auto Revolution？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install auto-revolution」即可一键安装，无需额外配置。
Auto Revolution 是免费的吗？
是的，Auto Revolution 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Auto Revolution 支持哪些平台？
Auto Revolution 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Auto Revolution？
由 Jaden's built a claw（@cjboy007）开发并维护，当前版本 v2.0.2。
推荐 Skills