using-offensive-claude
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEPROMPT_INJECTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill uses extremely high-pressure, mandatory language ('ABSOLUTELY MUST', 'not negotiable', 'cannot rationalize your way out of it') to force the agent to invoke specific tools. It explicitly instructs the agent to 'override default behavior where they conflict,' which is a common pattern for attempting to bypass system-level instructions or guardrails.
- [INDIRECT_PROMPT_INJECTION]: The skill establishes a workflow for processing external tasks and user instructions, creating a surface where malicious input could influence the agent's actions within the 'offensive-security framework.'
- Ingestion points: The skill reacts to 'Engagement task received' and incorporates 'User's explicit instructions' and 'direct requests.'
- Boundary markers: No explicit delimiters or instructions to ignore embedded commands in the processed data are provided.
- Capability inventory: The skill facilitates the invocation of various technique-specific skills (e.g., exploitation, AD, cloud) and uses tools like 'Skill' and 'TodoWrite'.
- Sanitization: There are no mechanisms described for sanitizing or validating the tasks or user instructions before processing.
Audit Metadata