The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest context from project-specific configuration files and existing specification artifacts to guide the exploration process. This ingestion of untrusted data from the local codebase creates a surface for indirect prompt injection.
Ingestion points: Project configuration files (config.yaml, config.yml) and existing specification files (e.g., spec.md, design.md, proposal.md) identified via the openspec status command.
Boundary markers: The instructions ground thinking in these files but do not specify clear delimiters or warnings to ignore instructions embedded within that data.
Capability inventory: The skill uses the openspec CLI via the Bash tool to read status, view artifacts, and scaffold new change documentation.
Sanitization: No explicit sanitization or content validation of the ingested files is described in the skill instructions.

openspec-explore