The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted input data, such as playtest notes and bug reports, to generate coordination artifacts, creating a surface for indirect prompt injection.\n- Ingestion points: The intake brief process and operating modes described in SKILL.md and references/operating-modes.md.\n- Boundary markers: The agent is instructed to normalize data into YAML blocks, providing basic structural separation.\n- Capability inventory: The skill has access to tools including Bash, Read, Write, Grep, and Glob.\n- Sanitization: There are no explicit instructions or mechanisms provided to sanitize ingested text or to ignore potential instructions embedded within external feedback or documents.

bmad-gds