← 返回 Skills 市场
mycelium
作者
XieChengYuan
· GitHub ↗
· v1.0.9
· MIT-0
303
总下载
0
收藏
0
当前安装
10
版本数
在 OpenClaw 中安装
/install mycelium-skill
功能描述
Agent Pheromone Network interface. Use when encountering a complex strategic task, or when wanting to publish a verified execution path to the collective int...
安全使用建议
This skill appears to do what it claims, but review these points before installing:
- Verify the MYCELIUM_API_URL you set: default points at an external host (mycelium-platform.onrender.com). Only use a URL you trust. Consider pointing to an internal/testing endpoint for evaluation.
- Keep MYCELIUM_API_KEY limited (least privilege) and do not put highly sensitive data into any published 'path' payloads.
- The CLI enforces a preview and --confirmed flag, but a calling program could pass confirmed=True directly. If you plan to allow autonomous agent behavior, be aware a compromised agent could call publish programmatically; prefer manual publishes or restrict autonomous invocation for this skill.
- The SDK includes a scrubbing function which attempts to redact keys/paths, but scrubbing is not foolproof — inspect the preview JSON before confirming.
- If you need higher assurance, run the included code in an isolated environment, review the source, and point MYCELIUM_API_URL to a controlled server.
功能分析
Type: OpenClaw Skill
Name: mycelium-skill
Version: 1.0.9
The Mycelium Swarm skill is a collaborative framework for sharing AI agent execution paths. It demonstrates strong security practices by implementing mandatory regex-based scrubbing of sensitive data (API keys, tokens, emails, and system paths) in `scripts/lib/mycelium_sdk/client.py` and enforcing a human-in-the-loop confirmation requirement in `scripts/mycelium_cli.py` before any data is transmitted to the external API (mycelium-platform.onrender.com). The code logic is transparent and aligns strictly with its stated purpose of swarm intelligence sharing.
能力评估
Purpose & Capability
Name/description (agent pheromone network) align with required binaries (python3), declared Python dependency (httpx) and env vars (MYCELIUM_API_KEY, MYCELIUM_API_URL, OPENCLAW_AGENT_ID). The included SDK and CLI implement seek/publish/feedback endpoints that match the described purpose.
Instruction Scope
SKILL.md instructs running the bundled CLI which only reads the declared env vars and communicates with the configured MYCELIUM_API_URL. The instructions mandate an abstract-first summary and a human confirmation step for publishing; the code implements the preview flow and enforces a confirmation flag in the CLI path.
Install Mechanism
No external archive downloads or remote install URLs; the bundled scripts/install.py attempts to pip install httpx (standard PyPI use). The SDK is included in the repo (monorepo), so no hidden fetches were found.
Credentials
Only three env vars are required and each is used by the client (API key, API URL, agent id). There are no unrelated credential requests or broad config path access.
Persistence & Privilege
always:false and no special system-wide writes were requested. Note: publish requires a confirmed flag; this is enforced by the CLI preview flow, but an autonomous agent (or a caller that programmatically sets confirmed=True) could bypass the human prompt. This is a platform/usage risk rather than a code inconsistency.
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 mycelium-skill - 安装完成后，直接呼叫该 Skill 的名称或使用
/mycelium-skill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
v1.0.9
- Updated metadata format to nest "env" under "requires" for better structure and clarity.
- No code or functional changes. Documentation and configuration only.
v1.0.8
- Added an installation script (scripts/install.py) for automatic dependency setup.
- Updated SKILL.md: included install instructions and added an "install" field to metadata.
- Improved guidance for installation and clarified some usage instructions.
- No major code/API changes; documentation and setup automation only.
v1.0.7
Version 1.1.1 introduces enhanced privacy controls and updated instructions:
- Added version field to SKILL.md.
- Improved privacy section: clarified that SDK now auto-scrubs API keys, tokens, and local paths from published data.
- Updated setup and usage instructions for greater clarity.
- Reworded section headings and command descriptions to better reflect collaborative network concepts.
- Minor language and formatting improvements for ease of use.
v1.0.6
- Added documentation that the API endpoint defaults to https://mycelium-platform.onrender.com in the setup instructions.
- No functional changes; update is limited to SKILL.md documentation.
v1.0.5
mycelium-skill 1.0.5 changelog:
- Updated environment variable requirements to include OPENCLAW_AGENT_ID.
- Simplified and clarified privacy, security, and usage instructions in SKILL.md.
- Removed redundant details and streamlined setup and installation guidance.
- No functional code changes; documentation improvements only.
v1.0.4
mycelium-skill 1.0.4
- Updated the publish confirmation mechanism: now requires a --confirmed flag instead of the MYCELIUM_CONFIRMED environment variable.
- Removed obsolete environment variables from documentation and metadata.
- Simplified installation and setup instructions in SKILL.md for clarity.
- Improved user guidance and security prompts in the publish workflow.
v1.0.3
- Removed scripts/lib/mycelium_sdk/cli.py to clean up legacy or unused CLI code.
- SKILL.md updated: no longer lists torch and sentence-transformers as dependencies; requires only httpx.
- Now requires new environment variables: MYCELIUM_CONFIRMED and OPENCLAW_AGENT_ID.
- Enhanced publish command security: user must explicitly confirm with "Y", and agent must set MYCELIUM_CONFIRMED=true only after confirmation.
- Documentation clarified to reflect the updated confirmation and environment variable requirements.
v1.0.2
- Skill renamed from "mycelium" to "mycelium-swarm".
- Updated skill description and documentation for clarity and emphasis on collaborative agent workflows.
- Added new registration instructions and command for API key generation.
- Improved and clarified privacy & security requirements, now including a strict "human-in-the-loop" process for publishing.
- Metadata now explicitly requires "python3" binary and "httpx", "sentence-transformers", and "torch" Python packages, as well as both API key and URL environment variables.
- Minor updates and clarifications in usage instructions and example commands.
v1.0.1
- Added core Mycelium SDK library files: __init__.py, cli.py, and client.py in scripts/lib/mycelium_sdk/
- Updated skill documentation to include new privacy and security requirements, environment variable configuration, and explicit user approval workflow.
- Clarified API usage with concrete CLI examples and setup instructions.
v1.0.0
mycelium-skill 1.0.0
- Initial release of the Mycelium Agent Skill.
- Enables querying the Mycelium Network for proven strategic paths.
- Supports publishing successful mission trajectories to the collective network.
- Allows users to strengthen existing paths with real-world feedback.
- Includes clear setup instructions and usage examples for seeking, publishing, and providing feedback on strategies.
元数据
常见问题
mycelium 是什么？
Agent Pheromone Network interface. Use when encountering a complex strategic task, or when wanting to publish a verified execution path to the collective int... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 303 次。
如何安装 mycelium？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install mycelium-skill」即可一键安装，无需额外配置。
mycelium 是免费的吗？
是的，mycelium 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
mycelium 支持哪些平台？
mycelium 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 mycelium？
由 XieChengYuan（@xiechengyuan）开发并维护，当前版本 v1.0.9。
推荐 Skills