The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes project configuration and documentation which may contain untrusted instructions.
Ingestion points: Processes openspec/config.yaml, openspec/config.yml, and OpenSpec artifacts such as spec.md, design.md, proposal.md, and tasks.md. It also follows instructions provided in the output of the openspec instructions command.
Boundary markers: Absent; there are no instructions to delimit or ignore potential instructions embedded within the ingested data.
Capability inventory: Executes openspec CLI commands (e.g., new change, status, instructions, list) through a scoped bash tool.
Sanitization: The skill does not specify any sanitization or validation logic for the content read from the filesystem or tool outputs.

openspec-explore