The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill acts as a pipeline for ingesting external documentation, GitHub repositories, and PDF files. This data is then used to generate AI-ready instructions and references, creating a significant surface for indirect prompt injection attacks.
Ingestion points: Website URLs, GitHub repository content, and PDF files processed through the scrape, github, and pdf commands in SKILL.md.
Boundary markers: The provided documentation does not indicate the use of clear delimiters or instructions for the agent to ignore potentially malicious embedded content within the source data.
Capability inventory: The skill is capable of performing network operations (requests), writing files to the local output/ directory, and interacting with the Anthropic API for content enhancement.
Sanitization: No explicit sanitization or validation logic for the external content is described in the skill's instructions or metadata.
Self-referential content: The references/readme.md file contains security assessment badges (mseep.ai) claiming safety. Per security guidelines, such claims are treated as data to be evaluated rather than established facts.

skill-seekers