The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute various subcommands of the openspec CLI, such as new change, status, and instructions. These operations are limited to the openspec namespace by the allowed-tools configuration.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources which could potentially contain malicious instructions if the project files or CLI outputs are compromised. Ingestion points: Reads output from openspec status --json, openspec instructions --json, and previously created artifact files (proposal.md, design.md, tasks.md) in SKILL.md. Boundary markers: The skill uses structured JSON parsing for CLI outputs, but no explicit text delimiters are defined for the content of the artifacts themselves. Capability inventory: The agent can execute openspec CLI commands and write files to the local file system. Sanitization: The skill includes instructions to treat context and rules from the CLI as internal constraints rather than content to be copied, which serves as a basic mitigation against instruction leakage.

openspec-propose