The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local Python scripts to manage campaign tracking and benchmark data. Evidence includes calls to campaign-tracker.py and benchmark_book.py in SKILL.md and marketing-ops.md.
[INDIRECT_PROMPT_INJECTION]: The skill ingests brand profiles and guidelines from the user's home directory (~/.claude-marketing/brands/), which could serve as a vector for malicious instructions to influence agent behavior.
Ingestion points: Files located at ~/.claude-marketing/brands/{slug}/profile.json and ~/.claude-marketing/brands/{slug}/guidelines/_manifest.json, plus reference files in skills/context-engine/.
Boundary markers: The instructions do not define delimiters or specific "ignore instructions" warnings for these data files.
Capability inventory: Execution of local scripts (campaign-tracker.py, benchmark_book.py) and ability to produce marketing design documents.
Sanitization: There is no mention of sanitization or schema validation for the ingested JSON content.
[DYNAMIC_EXECUTION]: The skill dynamically resolves the location of its scripts using the CLAUDE_PLUGIN_ROOT environment variable at runtime, which is used to construct paths for script execution.

marketing-automation