The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted external data, specifically job descriptions (JD) and resumes, which introduces an attack surface for indirect prompt injection.
Ingestion points: prompts/jd-driven-prep.md (Step 1 and Step 2) instructs the agent to ingest the full text of JDs and resumes.
Boundary markers: The instructions do not specify the use of clear delimiters or boundary markers to separate the untrusted input from the agent's core instructions.
Capability inventory: The skill has the capability to write HTML reports to the user's local file system at the path ~/Desktop/Claude skills/ (specified in prompts/jd-driven-prep.md, Step 7).
Sanitization: The instructions do not include requirements for the agent to sanitize or escape data extracted from external documents before it is processed or written to files.
[COMMAND_EXECUTION]: The skill instructions direct the agent to interact with the local file system by creating and writing HTML reports.
Evidence: prompts/jd-driven-prep.md Step 7 defines a specific path on the user's desktop (~/Desktop/Claude skills/bq-prep-<company>-<role>-<YYYYMM>.html) for outputting prep reports. While this is a documented feature of the skill, file system write access is a sensitive capability.

bq-skill