The Agent Skills Directory

[SAFE]: The skill operates within its defined scope as a specialist for statistical method selection, simulation design, and validation protocols, providing structured templates for coordination between different agent roles.
[EXTERNAL_DOWNLOADS]: The skill references industry-standard Python libraries such as SciPy, Statsmodels, PyMC, Stan, ArviZ, and NumPy. These are trusted, well-known resources for statistical computing and their use is consistent with the skill's primary function.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from user-supplied contexts and request IDs which are then used in logic and file operations. Evidence: 1. Ingestion points: The stats_request YAML structure in SKILL.md provides entry points for external data via id and context fields. 2. Boundary markers: No explicit delimiters or instructions are present to prevent the agent from obeying instructions embedded within the problem_statement or context. 3. Capability inventory: The skill utilizes a Write tool to generate files, which is a significant capability. 4. Sanitization: The skill instructions do not specify sanitization or validation for the request-id, creating a potential path traversal surface when generating progress reports at /tmp/progress-{request-id}.md.

statistician