The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data through the $ARGUMENTS placeholder in both SKILL.md and commands/swot.md without employing boundary markers or sanitization.
Ingestion points: User input is ingested via the $ARGUMENTS parameter which is appended to the instructions in SKILL.md and commands/swot.md.
Boundary markers: None present. The skill does not use delimiters (like XML tags or triple quotes) or explicit instructions to ignore commands embedded within the user-provided analysis subject.
Capability inventory: As the skill does not restrict tool access in its frontmatter, it retains access to the agent's default capabilities, which typically include filesystem access, shell execution, and network operations.
Sanitization: There is no evidence of input validation, escaping, or filtering for the data passed into $ARGUMENTS.

swot