The Agent Skills Directory

[COMMAND_EXECUTION]: The skill initiates the execution of a local script, revenue-simulator.py, to run its core simulation logic. This command execution is scoped to a specific provided script and handles local data processing.
[SAFE]: The skill reads from application-specific paths such as ~/.claude-marketing/brands/ and ~/.claude-marketing/sops/. These operations are limited to the skill's own data environment and do not involve sensitive system files like SSH keys, AWS credentials, or global environment variables.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes untrusted data from brand profiles and guidelines.
- Ingestion points: ~/.claude-marketing/brands/{slug}/profile.json and _manifest.json
- Boundary markers: Not specified in instructions
- Capability inventory: Execution of revenue-simulator.py
- Sanitization: No explicit validation mentioned
  While this surface exists, the risk is assessed as safe because the data is primarily used as numerical parameters for a mathematical simulation engine, which limits the potential for executing arbitrary agent instructions.

simulate