The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external project documents which could contain malicious instructions meant to subvert the agent's logic.\n
Ingestion points: The workflows require the agent to read various project files, including DESIGN.md, ENG.md, plan files (e.g., docs/plans/), and ticket files in docs/tickets/.\n
Boundary markers: There are no instructions to use specific boundary markers or delimiters (like XML tags or clear separators) to isolate external content from the system prompt when reading files.\n
Capability inventory: The skill uses Bash, Write, and Edit tools, which provide the capability to modify the filesystem or execute shell commands if prompted by injected content.\n
Sanitization: The workflow does not include steps for sanitizing, validating, or filtering the content of the read files before processing them.

eng-architect