The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local openspec CLI commands (new change, status, instructions) to orchestrate project changes. These operations are consistent with the skill's stated purpose and use the developer's own toolset.
[DATA_EXFILTRATION]: File operations are localized to the project's openspec/changes/ directory. The skill does not access sensitive system files (e.g., SSH keys, AWS credentials) or perform unauthorized network transfers.
[PROMPT_INJECTION]: Instructions such as "IMPORTANT" and "Guardrails" are used to define the agent's operational logic and output formatting. No attempts to override safety protocols or extract system prompts were detected.

openspec-propose