The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest and analyze untrusted external data provided in SBOM and VEX formats, which could potentially contain embedded instructions intended to manipulate the agent.
Ingestion points: The skill explicitly processes user-provided or environment-discovered SBOM file(s) and VEX document(s) as detailed in the 'Context the Agent Needs' section of SKILL.md.
Boundary markers: The skill includes a dedicated 'Prompt Injection Safety Notice' section that provides clear instructions to the agent to ignore any commands or instructions found within the data files, acting as a defensive boundary.
Capability inventory: The skill is configured with restricted tools (Read, Grep, Glob) for inspection only, and lacks high-risk capabilities like network access or arbitrary command execution.
Sanitization: While the skill describes structural validation processes (Step 1), it relies on natural language instructions rather than automated sanitization to handle potential injection attempts in the data fields.

sbom-analysis