The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill exposes a surface for indirect prompt injection as it is designed to ingest and process untrusted data from user-provided files such as stack.json, campaigns.json, and leads.json.
Ingestion points: Untrusted data enters the agent context through the JSON files passed to the local auditing and simulation scripts.
Boundary markers: No specific boundary markers or "ignore embedded instructions" warnings are defined in the instructions for the data being processed.
Capability inventory: The skill possesses the capability to execute local scripts and route questions to other specialized marketing skills. It does not exhibit network exfiltration or sensitive file access capabilities.
Sanitization: The data is parsed as structured JSON, which provides schema constraints but does not sanitize natural language values for potential instructional content.
[COMMAND_EXECUTION]: The skill uses the python command to run internal utility scripts (martech_stack_auditor.py, campaign_budget_allocator.py, lead_scoring_simulator.py). These scripts are self-contained, use standard libraries only, and are strictly limited to performing marketing-related calculations and audits based on the provided input files.

marketing-ops