context-builder
Generate LLM-optimized codebase context from any directory using context-builder CLI
Why use this skill?
Generate LLM-optimized markdown snapshots of any codebase. Featuring AST-aware signatures, smart filtering, and token-efficient formatting for deep AI analysis.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/igorls/context-builderWhat This Skill Does
The context-builder skill is an advanced agentic tool designed to transform complex, multi-file codebases into a single, LLM-optimized markdown document. By leveraging AST (Abstract Syntax Tree) awareness and smart heuristic sorting, the tool extracts the most critical structural information—such as function signatures and class definitions—while minimizing token consumption. It intelligently filters out noise by respecting .gitignore files and automatically excluding heavy directories like node_modules or binary assets, ensuring the output is perfectly formatted for AI analysis.
Installation
To integrate this skill into your environment, use the OpenClaw hub CLI:
clawhub install openclaw/skills/skills/igorls/context-builder
Ensure you have the Rust toolchain installed, as the underlying utility is a high-performance Rust binary. You can build it from source or use the verified binaries available on GitHub to ensure integrity. Verify installation by running context-builder --version in your terminal.
Use Cases
This skill is indispensable for developers who need to perform deep architectural analysis, cross-file bug hunting, or rapid project onboarding. It is particularly useful for agents tasked with documentation generation or those needing a consolidated 'snapshot' of a repository to feed into a large context window. It excels at summarizing large projects where direct file-by-file inspection would exceed standard token limits or cause cognitive drift.
Example Prompts
- "Generate a comprehensive context file for the project located at /home/user/src/api-server, including only .go and .yaml files, and save it to /home/user/tmp/context.md."
- "Create an AST-only signature report of the entire repository at /var/www/my-app to help me understand the function dependencies before I start refactoring."
- "Summarize the codebase at /home/user/workspace/web-ui, excluding the 'tests' directory, and output the result so I can ask the agent about the core state management logic."
Tips & Limitations
Always use the -f flag to restrict file types; this dramatically reduces the output size and improves the quality of the LLM's understanding. Avoid running this on root directories or sensitive system paths. The tool includes built-in protections against self-ingestion and binary file processing, but always review the generated file for sensitive credentials (API keys, secrets) before passing the content to a cloud-based LLM. Remember that the quality of the context is highly dependent on your inclusion/exclusion filtering strategy.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-igorls-context-builder": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: file-read, file-write