The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to run CLI commands via a project-local runner script located at .gitnexus/run.cjs. These commands manage repository analysis and indexing.
[EXTERNAL_DOWNLOADS]: The documentation describes using npx, bunx, or pnpm dlx to download and execute the gitnexus package and its dependencies from public registries to set up the environment.
[DATA_EXFILTRATION]: The wiki command can export repository metadata and generated documentation to external LLM providers and GitHub Gists. It also mentions storing API keys in ~/.gitnexus/config.json for persistence across sessions.
[INDIRECT_PROMPT_INJECTION]: The tool builds a knowledge graph by parsing source files in the repository. This graph is used as a data source for LLM documentation generation, creating a surface where malicious code comments could potentially influence the generated wiki content.
Ingestion points: Source files throughout the project are parsed by the analyze command as described in SKILL.md.
Boundary markers: No specific delimiters are mentioned to separate code content from instructions during the parsing or documentation generation phases.
Capability inventory: Includes local script execution (node), network-based LLM requests, and GitHub Gist publishing.
Sanitization: The documentation does not specify any sanitization or filtering of the source code before it is processed into the knowledge graph.

gitnexus-cli