The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface where untrusted input from the user or external knowledge base tools could influence agent behavior.
Ingestion points: The skill ingests data from the $ARGUMENTS variable (user-provided dilemma) and retrieved content from the mcp__solograph__kb_search tool.
Boundary markers: There are no explicit delimiters (such as XML tags or markdown fences) or 'ignore' instructions used when interpolating this external data into the agent's interactive reasoning process.
Capability inventory: The skill utilizes the Write tool to persist decision summaries to the file system (specifically the docs/decisions/ directory) and has access to Read, Grep, and Glob tools.
Sanitization: The instructions do not specify any validation or sanitization steps for the external content before it is included in the synthesized verdict or written to local files.

solo-stream