The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from local configuration files to determine brand voice and compliance guidelines.
Ingestion points: The skill reads multiple files from the user's home directory in ~/.claude-marketing/, including _active-brand.json, profile.json, _manifest.json, and various files within the sops/ directory.
Boundary markers: There are no explicit instructions or delimiters used to isolate the content of these files from the agent's primary instructions, meaning instructions inside a brand profile could influence agent behavior.
Capability inventory: The skill performs file read operations on the local file system and invokes external agents (content-creator, seo-specialist) to process the data.
Sanitization: No sanitization or validation logic is present to filter malicious instructions that might be embedded in the brand profiles or SOP documents.

content-brief