The Agent Skills Directory

[SAFE]: No malicious patterns detected. The skill implements a playbook interpreter that manages task dependencies, parallel execution, and success verification via tool usage or human confirmation.
[COMMAND_EXECUTION]: The skill facilitates command execution through the checks mechanism (type: command). This is a documented and intended feature for verifying task success (e.g., running tests) and follows standard developer workflows.
[PROMPT_INJECTION]: Instructions include clear directives for the agent to resolve !include tokens and substitute variables before execution. There are no attempts to bypass safety filters or override system constraints.
[DATA_EXFILTRATION]: No evidence of unauthorized data access or external transmission. Network operations are not requested, and file access is scoped to reading playbook files and verifying local artifacts.
[EXTERNAL_DOWNLOADS]: The skill does not perform remote downloads or reference external code repositories.

playbook-execute