The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read codebase files and OpenSpec artifacts (such as designs and proposals) to gather context. This ingestion of external, potentially untrusted data constitutes a surface for indirect prompt injection. However, the risk is mitigated by explicit guardrails within the skill that prevent the agent from implementing code or executing arbitrary commands outside the scoped CLI.
[COMMAND_EXECUTION]: The skill utilizes the openspec CLI via Bash to perform operations like openspec list, openspec status, and openspec store list. These commands are restricted to the openspec namespace as defined in the allowed-tools configuration, ensuring the agent operates within its intended scope.

openspec-explore