The Agent Skills Directory

[DATA_EXFILTRATION]: The skill accesses and manages highly sensitive directories including ~/.claude-marketing/credentials/ and ~/.claude-marketing/brands/. These locations are documented to contain API keys, OAuth tokens, and environment variables for multiple marketing platforms.
[COMMAND_EXECUTION]: The skill invokes local Python scripts, specifically credential-manager.py and execution-tracker.py, with various arguments to perform administrative tasks such as switching profiles and validating credentials.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting and acting upon data from local configuration files.
Ingestion points: The agent reads brand profiles and credential metadata from profile.json, _active-brand.json, and _active-profile.json to determine its execution flow.
Boundary markers: No explicit delimiters or instructions to ignore instructions embedded within the JSON configuration data are present.
Capability inventory: The skill possesses significant capabilities including the execution of local scripts via shell and full file system access to the application's configuration directory.
Sanitization: There is no evidence of sanitization or schema validation for the data read from these files before it is processed by the agent or passed to the execution scripts.

credential-switch