The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external data that could contain instructions designed to manipulate the agent's behavior.
Ingestion points: The skill reads project-specific files listed in the contextFiles output (such as proposals, specs, designs, and task lists) and incorporates context and operationGuidance fields returned by the openspec instructions CLI command.
Boundary markers: The instructions explicitly tell the agent to treat these inputs as prompt-level guidance, and it notes that if the context conflicts with built-in instructions, the agent should report the conflict and preserve the controlling value.
Capability inventory: The skill uses Bash(openspec:*) to interact with the environment and is designed to modify the local codebase to implement tasks.
Sanitization: The skill lacks explicit sanitization or filtering for the external content it reads, relying instead on the agent's ability to identify and report conflicting instructions.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute openspec CLI commands (status, list, instructions, etc.) to drive the implementation workflow.

openspec-apply-change