A specialized tool for understanding and summarizing local PDF, video, and audio files using the Google Gemini API.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install gemini-reader
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install gemini-reader using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Gemini Reader?
Gemini Reader is a high-performance extension designed for the Openclaw Skills framework, enabling AI agents to process non-text assets that fall outside standard LLM context windows. By utilizing the Gemini Python SDK, this skill allows users to extract intelligence from complex PDF documents, video files, and audio recordings. It serves as a bridge for developers who need to incorporate deep multimodal analysis into their automated workflows without manual data entry.
This skill is particularly valuable for users of Openclaw Skills who require robust transcription and summarization capabilities for media. It handles the heavy lifting of file uploading, processing via Google's latest models like Gemini 2.5 and 3.1, and ensuring temporary files are cleaned up post-analysis for maximum efficiency.
Gemini Reader Use Cases
- Summarizing dense technical documentation and research papers in PDF format.
- Generating verbatim transcriptions and key takeaways from audio recordings like meetings or interviews.
- Analyzing video lectures or presentations to identify specific topics and visual content.
- Extracting data tables and structured information from non-searchable document formats.
- Automating the processing of media libraries for content indexing and archival.
How Gemini Reader Works
- The user provides a path to a local PDF, video, or audio file along with a descriptive prompt.
- The skill performs a security check, validating the file extension and ensuring the path is safe to access.
- The file is uploaded to the Google Gemini API using the File Upload API, which supports large media files.
- The specified Gemini model processes the file based on the user's prompt (e.g., summarization or transcription).
- The AI-generated analysis is returned to the user or saved to a specified output file.
- The skill triggers a cleanup process to ensure the remote file is deleted from the Gemini server after the session.
Gemini Reader Setup
To integrate this capability into your Openclaw Skills environment, ensure you have the required Python SDK and API credentials configured.
# Install the Gemini Python SDK
pip install google-genai
# Set your API key as an environment variable
export GEMINI_API_KEY='your_gemini_api_key_here'
Gemini Reader Data Schema & Taxonomy
Gemini Reader supports a wide range of media formats and organizes data through a standardized processing pipeline.
| Media Type | Supported Extensions | Common Metadata Handled |
|---|---|---|
| Documents | Page count, text content, embedded tables | |
| Video | .mp4, .webm, .mov, .avi, .mkv | Visual cues, spoken dialogue, timestamps |
| Audio | .mp3, .wav, .m4a, .ogg | Speech-to-text, speaker tone, audio duration |
Output can be directed to the terminal or a persistent text file using the --output flag.
Gemini Reader Advanced Features
- Support for multiple model aliases including 3-flash for speed and 3.1-pro for complex reasoning.
- Automated MIME type detection ensures seamless handling of diverse file formats without manual configuration.
- Direct API integration allows for high-throughput processing without the limitations of standard CLI sandboxes.
- Integrated security layers that prevent the analysis of sensitive system paths or symlinks.
- Flexible deployment options that complement other Openclaw Skills for end-to-end automation pipelines.
SKILL.md
Loading
Related Openclaw Skills
Web Search
A powerful, API-free internet search tool for AI agents to retrieve real-time data from Baidu and Bing.
DocStrange (Nanonets Document Extraction)
DocStrange is a powerful document extraction API by Nanonets that converts PDFs and images into structured Markdown, JSON, or CSV data with field-level confidence scoring.
Pocket TTS: Local Offline Text-to-Speech
A fully local, CPU-based text-to-speech engine providing high-quality audio generation and voice cloning without internet dependencies.
deploy-agent
A multi-step deployment agent that orchestrates the workflow from local build and testing to GitHub hosting and Cloudflare Pages deployment.
Vibe Coding Workflow
A structured 5-phase development framework designed to guide AI agents from vague requirements to production-ready code and iteration.
Vibe Coding Workflow
A systematic 5-phase framework designed to transform vague ideas into production-ready software through structured AI collaboration.
METADATA
Featured*
