The Agent Skills Directory

[SAFE]: No security issues were detected during the analysis. The skill follows best practices for a documentation generator and limits its operations to the local project environment.
[COMMAND_EXECUTION]: The skill utilizes a local Python script, generate_llms_manifest.py, to perform a deterministic crawl of the repository. This script identifies documentation files and extracts a summary from the README.md. The execution is scoped to the project root and does not involve shell injection or remote commands.
[PROMPT_INJECTION]: The skill ingests data from the local project environment (file paths and README content) which acts as an attack surface for indirect prompt injection. This surface is inherent to the skill's purpose of summarizing project documentation.
Ingestion points: The generate_llms_manifest.py script processes all file paths in the provided root directory and reads the first paragraph of the local README.md file.
Boundary markers: The skill does not implement specific delimiters or 'ignore' instructions to separate ingested content from system instructions during the drafting phase.
Capability inventory: The tool facilitates reading local files and drafting new documentation files based on those contents.
Sanitization: No content sanitization or filtering of the ingested project data is performed before it is presented to the agent.

llms-txt