The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from repository files such as quest/AGENTS.md and existing quest/questline documents. This provides a surface for indirect prompt injection where malicious instructions embedded in those files could influence the agent's planning logic.
Ingestion points: SKILL.md instructs the agent to read quest/AGENTS.md and search other quests and questlines.
Capability inventory: The skill can create, update, or delete files, perform git commits, and open pull requests, as noted in SKILL.md.
Boundary markers: The skill lacks explicit instructions for using delimiters or boundary markers when processing these files.
Sanitization: No specific sanitization or filtering logic is mentioned for the content retrieved from the repository files.

plan-quest