The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute openspec CLI commands. These commands are used to discover project roots, load context, create new change sets, and retrieve instructions for artifact generation.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources that could contain malicious instructions.
Ingestion points: The agent reads the context field from openspec/config.yaml (or config.yml), processes user-provided change descriptions, and reads existing artifact files from the project directory.
Boundary markers: The instructions include a specific "Project check" and "Planning boundary" intended to prevent the agent from being diverted from its workflow, though no formal data delimiters are used during interpolation.
Capability inventory: The skill has the ability to execute shell commands via the openspec CLI and write new files to the local filesystem.
Sanitization: There is no evidence of content sanitization or validation of the ingested project data before it is incorporated into the agent's decision-making process.

openspec-propose