← 返回 Skills 市场
381
总下载
0
收藏
2
当前安装
7
版本数
在 OpenClaw 中安装
/install workflow-tools
功能描述
Work smarter with loop detection, parallel decisions, and file size analysis
安全使用建议
This skill appears to do what it says (scan code/workspaces for markers, evaluate parallelization, and check file sizes). Before installing: 1) Inspect the two config files (.openclaw/workflow-tools.yaml and .claude/workflow-tools.yaml) for any sensitive values. 2) Avoid running `/wt` with no path (default=current) on directories that contain secrets—pass an explicit, narrow path instead. 3) Be cautious with `/wt subworkflow` and the --background option: spawning other ClawHub skills can cause those skills to use network access or credentials not mentioned here. Review and vet any skills you allow it to spawn (and their permissions). 4) Verify the upstream source/repo and dependency skills (failure-memory, constraint-engine, context-verifier) before installing. If you want to be extra cautious, run this skill in an isolated workspace or sandbox first to see what files it reads and what other skills it invokes.
功能分析
Type: OpenClaw Skill
Name: workflow-tools
Version: 1.5.1
The workflow-tools skill (v1.5.1) provides utilities for loop detection, parallelization analysis, and file size monitoring (MCE). The logic involves scanning local directories for patterns like 'TODO' or 'FIXME' and analyzing file lengths to suggest refactoring, with results stored in the local workspace. While the skill can read arbitrary user-specified paths and spawn other installed skills (subworkflows), these capabilities are transparently documented in SKILL.md and align with its stated purpose without evidence of malicious intent, data exfiltration, or unauthorized execution.
能力评估
Purpose & Capability
Name/description (workflow utilities: loop detection, parallel decisions, MCE analysis, subworkflows) align with declared config/output paths and the SKILL.md. Declared dependencies on other workflow-related skills are plausible for a consolidated tool.
Instruction Scope
SKILL.md explicitly reads user-specified directories/files (e.g., scans current working directory by default) which is consistent with file-analysis features, but it also supports 'subworkflow' spawning (task→spawn(clawhub.skill)) and background runs. The doc asserts 'No external APIs or third-party services are called,' yet spawning other ClawHub skills can cause those other skills to call external services or access credentials. This is a contradiction and expands runtime scope beyond what the skill's own manifest declares.
Install Mechanism
Instruction-only skill with no install spec or code files — lowest install risk. It lists other skills as dependencies to install via openclaw, but does not auto-download arbitrary binaries or archives.
Credentials
No environment variables or secrets are requested by the skill itself and only two config paths are declared (.openclaw/workflow-tools.yaml and .claude/workflow-tools.yaml). However, those YAML files (or subworkflows spawned) could contain or cause access to credentials—this is not surfaced by the skill manifest and should be reviewed before use.
Persistence & Privilege
always:false and user-invocable:true (default autonomous invocation allowed) — appropriate for this kind of tool. The skill does not request permanent presence or claim to modify other skills' configs, but it can spawn other skills and run background subworkflows which increases operational reach.
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 workflow-tools - 安装完成后，直接呼叫该 Skill 的名称或使用
/workflow-tools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
v1.5.1
- Updated author email in SKILL.md
- Bumped version from 1.4.0 to 1.5.0 in SKILL.md
- No logic or feature changes—documentation and metadata update only
v1.5.0
**Changelog for workflow-tools v1.5.0**
- Updated SKILL.md: Expanded and improved descriptions, especially around security and agent integration.
- Clarified data handling: Now explicitly states the skill uses the agent's configured model for workflow analysis (previously emphasized “instruction-only”).
- Adjusted tags to broaden discoverability (added automation, orchestration, decision-making, task-management).
- Minor clarifications and language updates throughout documentation for better user understanding.
- No changes to commands, arguments, or configuration structure.
v1.4.0
**Added explicit file access and security disclosure in documentation.**
- Clarified that skill reads user-specified files and directories for analysis via `/wt loops` and `/wt mce`.
- Added a warning section about file access and security considerations.
- No changes to commands, arguments, or logic.
- Updated documentation only; functionality unchanged.
v1.3.0
workflow-tools 1.3.0
- Version bump from 1.2.0 to 1.3.0 in SKILL.md.
- No functional or documentation changes beyond version update.
v1.2.0
**workflow-tools 1.2.0** — Now instruction-only, no model invocation
- Skill is now fully instruction-only (`disable-model-invocation: true`), providing frameworks and utilities without invoking AI models.
- All references to AI model invocation and data processing by external services have been removed or clarified.
- Data access is restricted to the workspace paths declared in skill metadata.
- Maintains all existing workflow analysis commands and configuration structures.
v1.1.0
workflow-tools 1.1.0
- Added a `metadata` section specifying required config and workspace paths, improving integration with OpenClaw and related tools.
- No changes to core features, logic, or usage instructions.
- Version updated to reflect metadata and structural enhancements.
v1.0.0
Initial release: Combines essential workflow utilities (loop detection, parallelization decisions, file size analysis, subworkflow spawning) into a unified skill.
- Detects open loops (DEFERRED, PLACEHOLDER, TODO markers) in code and project files.
- Offers a structured framework to decide between parallel and serial execution using five key factors.
- Performs file size analysis (Minimal Context Engineering, MCE), suggesting splits when thresholds are exceeded.
- Allows spawning of subworkflows to delegate tasks to other skills.
- Configuration supports multiple file formats and granular pattern/settings control.
- Removes redundant "pbd-strength-classifier" for streamlined workflow analysis.
元数据
常见问题
Workflow Tools 是什么？
Work smarter with loop detection, parallel decisions, and file size analysis. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 381 次。
如何安装 Workflow Tools？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install workflow-tools」即可一键安装，无需额外配置。
Workflow Tools 是免费的吗？
是的，Workflow Tools 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Workflow Tools 支持哪些平台？
Workflow Tools 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 Workflow Tools？
由 Lee Brown（@leegitw）开发并维护，当前版本 v1.5.1。
推荐 Skills