The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill represents a potential surface for indirect prompt injection as it processes untrusted problem statements and algorithmic requirements from external agents or users.
Ingestion points: Data enters through the math_request input structure, specifically the problem_statement and context fields defined in SKILL.md.
Boundary markers: The skill does not define specific delimiters or instructional barriers to isolate processed data from the agent's core instructions.
Capability inventory: The skill utilizes Write tools to generate design specifications and progress reports in the /tmp directory.
Sanitization: No explicit evidence of input validation, filtering, or sanitization was found in the provided skill definitions.
[EXTERNAL_DOWNLOADS]: The skill references standard, well-known libraries for numerical and mathematical computation.
Details: Mentions dependencies on numpy, scipy, scikit-learn (via sklearn.neighbors), Annoy, and Faiss. These are trusted packages from reputable organizations and are standard in mathematical software development.

mathematician