← Back to Skills Marketplace
focusshang

MinerU OCR Local & API

by ABear · GitHub ↗ · v1.1.4 · MIT-0
cross-platform ✓ Security Clean
225
Downloads
1
Stars
0
Active Installs
5
Versions
Install in OpenClaw
/install mineru-ocr-local-api
Description
Parse complex PDFs and document images with MinerU through either the hosted MinerU API or the local open-source MinerU runtime. Use when Codex, OpenClaw, Cl...
Usage Guidance
This skill appears to do what it says: it either uploads documents to the MinerU hosted API (requires MINERU_API_TOKEN) or runs a local MinerU CLI if you have it installed. Before installing, consider: (1) If you use API mode your document bytes (or URL content) will be sent to mineru.net — do not use API mode for highly sensitive documents unless you trust that service and token handling. (2) Local mode executes a mineru binary or python -m mineru.cli.client via subprocess; ensure you install and trust the upstream MinerU runtime if you want processing to stay local. (3) The tool saves JSON envelopes and extracted artifacts under the system temp directory by default — verify and remove those files if needed. (4) You will need Python and the httpx dependency available in the environment to run the wrapper. Otherwise the skill's requests and behavior are proportional to its stated purpose.
Capability Analysis
Type: OpenClaw Skill Name: mineru-ocr-local-api Version: 1.1.4 The skill bundle provides a legitimate interface for the MinerU OCR service, supporting both local CLI execution and the hosted API. The code in `scripts/lib.py` and `scripts/mineru_caller.py` uses standard libraries like `subprocess` and `httpx` to handle document parsing and file transfers, which are consistent with the stated purpose. No evidence of malicious intent, data exfiltration, or harmful prompt injection was found; the use of environment variables for API tokens and local paths is standard for this type of integration.
Capability Assessment
Purpose & Capability
Name/description match the code and instructions: the package is a wrapper that either calls the MinerU hosted API (needs MINERU_API_TOKEN) or invokes the local MinerU CLI. Required binaries (python) and the listed environment variables relate to MinerU operation. No unrelated credentials or unrelated system-wide accesses are requested.
Instruction Scope
Runtime instructions direct the agent to run python scripts/mineru_caller.py which (depending on mode) will upload local files to the MinerU hosted API, poll tasks, download result archives, or invoke a local mineru CLI via subprocess. The instructions and code reference only MinerU endpoints, temp directories, and local mineru configuration; they do not attempt to read unrelated system files. Note: in API mode user documents (local-file bytes or remote URLs) will be transmitted to the MinerU service, and downloaded artifacts are written under the temp root.
Install Mechanism
This is instruction-only (no automated installer). A requirements.txt and a pip dependency (httpx) are declared but no install script is present; requiring python and httpx is proportional for a Python CLI wrapper. No downloads from untrusted URLs or archive extraction are performed by the skill itself.
Credentials
The primary credential is MINERU_API_TOKEN which is appropriate and required only for the hosted API mode. Optional environment variables configure API base URL, timeouts, or local mineru invocation. No unrelated secrets or multiple unrelated service credentials are requested.
Persistence & Privilege
always is false; the skill does not request permanent/resident privileges or attempt to modify other skills or global agent configuration. Autonomous invocation is allowed but that is the default platform behavior and not by itself a concern.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install mineru-ocr-local-api
  3. After installation, invoke the skill by name or use /mineru-ocr-local-api
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.1.4
Improve missing-token guidance with explicit setup location and copy-paste PowerShell commands for API mode.
v1.1.3
Declare runtime metadata in SKILL.md frontmatter, remove user-specific validation path, and improve ClawHub security review accuracy.
v1.1.2
Remove auxiliary test scripts and tighten the published package to reduce false-positive security flags; no runtime behavior changes.
v1.1.1
Broaden skill description to support Codex, OpenClaw, Claude Code, and similar coding agents; no functional changes.
v1.1.0
Add local open-source MinerU mode alongside hosted API mode.
Metadata
Slug mineru-ocr-local-api
Version 1.1.4
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 5
Frequently Asked Questions

What is MinerU OCR Local & API?

Parse complex PDFs and document images with MinerU through either the hosted MinerU API or the local open-source MinerU runtime. Use when Codex, OpenClaw, Cl... It is an AI Agent Skill for Claude Code / OpenClaw, with 225 downloads so far.

How do I install MinerU OCR Local & API?

Run "/install mineru-ocr-local-api" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is MinerU OCR Local & API free?

Yes, MinerU OCR Local & API is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does MinerU OCR Local & API support?

MinerU OCR Local & API is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created MinerU OCR Local & API?

It is built and maintained by ABear (@focusshang); the current version is v1.1.4.

💬 Comments