concept-forge
Pass
Audited by Gen Agent Trust Hub on May 12, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill uses personas and instructional frameworks for concept refinement without attempting to bypass safety guardrails or extract system-level prompts.\n- [DATA_EXFILTRATION]: There are no network operations, external requests, or attempts to access sensitive local data such as credentials or SSH keys.\n- [REMOTE_CODE_EXECUTION]: No remote code downloads or execution patterns (e.g., curl|bash) were detected in the instructions or workflow.\n- [COMMAND_EXECUTION]: The skill does not utilize shell commands, subprocesses, or platform-level system operations.\n- [SAFE]: No obfuscation techniques, hidden characters, or hardcoded secrets were identified within the markdown content.
Audit Metadata