The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content from academic papers and GitHub repositories using the feynman alpha get and feynman alpha code commands. This external data is brought into the agent's context, creating a surface for indirect prompt injection where malicious instructions hidden in a paper's text or code could attempt to influence the agent's behavior.
Ingestion points: The feynman alpha get, feynman alpha ask, and feynman alpha code commands in SKILL.md ingest external paper content and source code.
Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands within the fetched academic content.
Capability inventory: The skill has the capability to execute shell commands (feynman alpha) and store persistent data (feynman alpha annotate).
Sanitization: No explicit sanitization or filtering of the external content is mentioned.
[COMMAND_EXECUTION]: The skill is designed to interact with the system via the feynman alpha command-line utility. It instructs the agent to execute shell commands to search, fetch, and annotate research papers.
[EXTERNAL_DOWNLOADS]: The skill retrieves data from external sources including the arXiv paper repository and GitHub code repositories. These references target well-known academic and development services.

alpha-research