← 返回 Skills 市场
174
总下载
0
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install moltbook-digest
功能描述
Collect Moltbook posts and comments, build an evidence pack, and interpret it through either the calling agent or LiteLLM.
安全使用建议
This skill appears coherent for Moltbook research, but review a few things before installing: 1) If you plan to enable the LiteLLM/provider path, only add API keys you trust and do so in a secure config or env var — supplying keys allows the script to make outbound calls. 2) 'uv sync' will install Python packages (litellm, pyyaml); consider using a virtualenv or isolated environment and pin package sources. 3) The script uses public Moltbook API endpoints by default (no auth required), but behavior may change if those endpoints are later restricted. 4) If you want to avoid any external network calls, run with --analysis-mode none (collection only) and keep provider configs empty. 5) If you are not comfortable providing provider credentials, audit the code paths that call litellm/providers (search for provider/api_key usage) prior to adding secrets. Overall the skill is consistent with its stated purpose.
功能分析
Type: OpenClaw Skill
Name: moltbook-digest
Version: 0.1.2
The bundle is a specialized research tool designed to aggregate and analyze discussions from Moltbook. The core logic in `scripts/moltbook_digest.py` uses standard API interactions to fetch posts and comments, while the `SKILL.md` and `config.example.yaml` provide structured instructions for the AI agent to generate research reports. There is no evidence of malicious intent, unauthorized data exfiltration, or harmful prompt injection; the instructions are strictly aligned with the stated purpose of evidence-based analysis and reporting.
能力评估
Purpose & Capability
Name/description match the code and instructions. The script queries Moltbook API endpoints (DEFAULT_BASE_URL), expands posts/comments, and writes evidence and report artifacts. The declared runtime binary 'uv' matches the SKILL.md examples. Declared project dependencies (litellm, pyyaml) align with the script's ability to call external LLM providers and parse YAML config.
Instruction Scope
SKILL.md limits actions to collection, expansion, and analysis steps and instructs the agent to read/write the generated files (digest.md, evidence.json, analysis_input.md, agent_handoff.md, analysis_report.md). It explicitly prefers the public API over scraping and warns not to reveal API keys. There are no instructions to read unrelated system files or to exfiltrate secrets.
Install Mechanism
No packaged install spec is included (instruction-only), but SKILL.md expects using the 'uv' tool to sync the project which will install dependencies from pyproject.toml (litellm, pyyaml). These are public PyPI packages — this is expected for LLM integration but does create a normal package-install footprint. No arbitrary download URLs or archive extraction are present.
Credentials
The skill requires no environment variables by default. config.example.yaml exposes optional provider API keys and lists common env var names (OPENAI_API_KEY, ANTHROPIC_API_KEY, etc.) for providers the script can call. This is proportionate to the optional 'LiteLLM' analysis path, but users should understand that supplying provider credentials enables outbound calls to external LLM APIs.
Persistence & Privilege
The skill is user-invocable, not forced-always, and does not request elevated system privileges. It writes output files into the project/baseDir (expected behavior). It does not modify other skills or system-wide agent settings.
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 moltbook-digest - 安装完成后，直接呼叫该 Skill 的名称或使用
/moltbook-digest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
v0.1.2
moltbook-digest 0.1.2
- Refactored documentation for clarity and conciseness, emphasizing evidence pack collection and reporting through agent or LiteLLM paths.
- Removed the sample OpenAI agent configuration file (agents/openai.yaml).
- Updated setup and config guidance to clarify provider use and key management.
- Described output contract with fixed filenames and clearer handoff expectations for agent and LiteLLM modes.
- Added guidance for handling vague user requests and documenting assumptions in the report.
v0.1.1
moltbook-digest 0.1.1
- SKILL.md significantly updated: added guidance for feed-based collection, monitoring scenarios, and continuous tracking via history directory.
- Digest output now includes a reader-facing brief section while retaining full post and comment data for analysis.
- Clarified and expanded instructions for interpreting mode selection and automated tracking of topic or submolt over time.
- Updated markdown output requirements for analysis (compact references, required sections, etc.).
v0.1.0
Initial release of moltbook-digest.
- Enables research and deep summarization of Moltbook posts and comment threads via Moltbook's public API.
- Supports keyword-based discovery, semantic search, trend analysis, and structured reporting on discussions.
- Outlines a recommended workflow: framing the research brief, evidence gathering, selective expansion, and rigorous analysis.
- Provides dual analysis methods: standalone LLM (LiteLLM) and agent-based interpretation.
- Designed for read-only research by default, with clear separation before any data write actions.
元数据
常见问题
MoltBook Digest 是什么？
Collect Moltbook posts and comments, build an evidence pack, and interpret it through either the calling agent or LiteLLM. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 174 次。
如何安装 MoltBook Digest？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install moltbook-digest」即可一键安装，无需额外配置。
MoltBook Digest 是免费的吗？
是的，MoltBook Digest 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
MoltBook Digest 支持哪些平台？
MoltBook Digest 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 MoltBook Digest？
由 Zhiwei Li（@mtics）开发并维护，当前版本 v0.1.2。
推荐 Skills