← 返回 Skills 市场
Meaningful Chunker
作者
DaymyanDogg
· GitHub ↗
· v1.2.2
· MIT-0
139
总下载
1
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install meaningful-chunker
功能描述
Graph-based code intelligence API. Query any indexed codebase for architecture understanding, debugging, refactor safety analysis, and design principle mappi...
安全使用建议
This skill appears internally consistent with a hosted code-analysis service. Before installing or using it: (1) Do not upload private/proprietary code to the public hosted endpoint unless you trust the operator or have a written agreement/privacy policy; scanning a repo or local path transmits code to CHUNKER_API_URL. (2) Prefer using a self-hosted instance if you need to analyze sensitive code. (3) Treat CHUNKER_API_KEY as a secret — use least-privilege or ephemeral keys if available and rotate them regularly. (4) Verify the CHUNKER_API_URL and registration endpoints (the README references a railway.app host) before entering keys, and test with non-sensitive public repos first. (5) Review the service's data retention and deletion policy so cloned/processed code is not stored longer than you expect.
功能分析
Type: OpenClaw Skill
Name: meaningful-chunker
Version: 1.2.2
The 'meaningful-chunker' skill is a documentation-based bundle that provides instructions for an AI agent to interact with a graph-based code intelligence API. It facilitates codebase scanning and natural-language querying for architectural analysis and debugging. The instructions in SKILL.md are transparent, well-structured, and align perfectly with the stated purpose. While the tool involves sending code to a third-party API (meaningful-chunker-production.up.railway.app), this is the core functionality of the service and is clearly documented. No evidence of malicious intent, credential theft, or prompt injection was found.
能力标签
能力评估
Purpose & Capability
Name/description match the declared requirements: it is an API-backed code-intelligence tool and it only asks for a base URL and API key (CHUNKER_API_URL, CHUNKER_API_KEY). No unrelated binaries, env vars, or installs are requested.
Instruction Scope
Instructions are focused on scanning a repo and querying the API. Important operational detail: POST /scan clones or uploads repositories (or accepts a local project_path for self-hosted instances), which effectively transmits repository contents to the remote service. This is expected for the stated purpose but is a privacy/data-exfiltration consideration you must accept before scanning private code.
Install Mechanism
No install spec and no code files — instruction-only skill. Nothing will be written to disk by the skill itself during install, which reduces supply-chain risk.
Credentials
Only two env vars are required (API base URL and API key); this is proportionate. CHUNKER_API_KEY is the primary credential and grants the ability to scan and query codebases on the remote host, so treat it as highly sensitive.
Persistence & Privilege
Skill is not always-enabled and uses normal model-invocation defaults. It does not request or imply modification of other skills or system-wide configuration.
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 meaningful-chunker - 安装完成后，直接呼叫该 Skill 的名称或使用
/meaningful-chunker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
v1.2.2
meaningful-chunker v1.2.2
- Updated environment variable documentation to specify free usage limits (100 queries/month, resets monthly) and Pro upgrade (2,000/month).
- Clarified that no credit card is required for a free API key.
- No functional or API changes. Documentation improvement only.
v1.2.1
- Updated setup instructions to include direct links for obtaining a free or Pro API key.
- Clarified that the API key is required and added details on where to register for access.
- No changes to version number or core functionality.
- Documentation improvements only; no code changes detected.
v1.2.0
- Major documentation update: SKILL.md now provides comprehensive setup instructions, usage guidelines, endpoint details, and response explanations.
- Added step-by-step quick start, authentication info, and environment variable requirements.
- Clarified how to scan codebases and check scan progress before querying.
- Detailed the intent and usage for all four query endpoints: architecture, debug, refactor, and philosophy.
- Included practical curl command examples for every workflow step.
- Introduced a Quick Read Guide and deeper explanation of response fields and context tiers for faster, more accurate results.
元数据
常见问题
Meaningful Chunker 是什么？
Graph-based code intelligence API. Query any indexed codebase for architecture understanding, debugging, refactor safety analysis, and design principle mappi... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 139 次。
如何安装 Meaningful Chunker？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install meaningful-chunker」即可一键安装，无需额外配置。
Meaningful Chunker 是免费的吗？
是的，Meaningful Chunker 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Meaningful Chunker 支持哪些平台？
Meaningful Chunker 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 Meaningful Chunker？
由 DaymyanDogg（@daymyandogg）开发并维护，当前版本 v1.2.2。
推荐 Skills