The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes system commands to determine the current state of the workspace. * Evidence: Uses git status -s, gh pr list, and gh issue list in SKILL.md to provide project context.
[INDIRECT_PROMPT_INJECTION]: The skill processes project-level files that could contain malicious instructions designed to influence the agent's behavior. * Ingestion points: Reads DATA/HANDOFF.md and DATA/buffer/session-*.md in SKILL.md. * Boundary markers: Absent; there are no specific markers or instructions provided to the agent to treat this ingested content as untrusted or to ignore instructions within them. * Capability inventory: The skill has access to shell command execution (git, gh) and file reading within the project directory. * Sanitization: Absent; the skill instructs the agent to distill and summarize these files directly into its context.

onboard