The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources and local files, which could contain malicious instructions designed to subvert the agent's behavior.
Ingestion points: Data is collected from external websites using WebFetch and WebSearch, and from local files using Glob and Read as specified in SKILL.md and references/edge-cases.md.
Boundary markers: The instructions do not define clear delimiters or include warnings to ignore instructions within the analyzed content.
Capability inventory: The skill possesses capabilities like Write, Edit, and Bash, which could be misused if the agent is successfully injected.
Sanitization: There are no documented procedures for sanitizing or validating external content before it is processed by the agent.

brand-voice-analyzer