The Agent Skills Directory

[SAFE]: The skill files consist of documentation and templates for business pricing models. No sensitive data access, credential harvesting, or network communication patterns were identified.- [DYNAMIC_EXECUTION]: The script scripts/validate.py performs local file system structure checks and parses configuration files. It uses yaml.safe_load() for YAML processing, which is a recommended security practice to prevent arbitrary code execution.- [INDIRECT_PROMPT_INJECTION]: 1. Ingestion points: The skill defines parameters for pricing_model and target_segment in SKILL.md. 2. Boundary markers: These inputs are restricted to specific allowed values via enums in the manifest. 3. Capability inventory: Analysis of scripts/validate.py shows no shell execution, file system writes, or network capabilities. 4. Sanitization: Input is validated against a whitelist of expected enum values.

pricing