sdd-riper-one
Pass
Audited by Gen Agent Trust Hub on May 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a structured engineering methodology that prioritizes human review and persistence of task state in local 'Spec' files. It enforces safety boundaries like 'No Spec, No Code' and 'No Approval, No Execute'.
- [COMMAND_EXECUTION]: The skill utilizes local Python utility scripts (
scripts/archive_builder.pyandscripts/default_prompt_check.py) to automate administrative tasks such as archiving task history and maintaining project configuration files (e.g.,AGENTS.md,.cursorrules). These scripts are self-contained and use standard library modules for local file operations. - [DATA_EXFILTRATION]: No network activity or external communication patterns were identified. All data processing and storage are confined to the local file system within the user's workspace.
- [PROMPT_INJECTION]: The skill includes strong instructional directives to ensure the agent follows the RIPER protocol (e.g., 'FORBIDDEN from writing code until...'). These are benign process-enforcement guidelines and do not attempt to override platform-level safety filters.
- [EXTERNAL_DOWNLOADS]: The skill does not perform any remote downloads or fetch code from external URLs. Dependencies used by the provided scripts are restricted to the Python standard library.
- [CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or unsafe handling of environment variables were detected.
Audit Metadata