skill-forge
Warn
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill includes a tool
skill_forge_createthat automates the creation of files within the.agents/skills/directory. This allows the agent to modify its own capabilities and persist new logic implementations on the filesystem. - [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection through the
objectiveparameter. Because this parameter is used to generate the logic and instructions for new skills, a crafted input could trick the agent into forging skills that contain malicious behavior or bypass safety constraints.
Audit Metadata