multi-touch-sequence
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to process conversation threads, which are external data sources that may contain malicious instructions meant to subvert the agent's behavior.
- Ingestion points: The prompt template in
SKILL.mdrequires the user to provide external data via[paste the last message I sent, or the full thread]. - Boundary markers: The prompt template lacks explicit delimiters or structural markers to separate the untrusted conversation data from the instructions, increasing the risk that the agent may follow instructions embedded within the lead's messages.
- Capability inventory: The skill is designed for text generation only and does not include any scripts, file system access, or network capabilities.
- Sanitization: No input validation or sanitization is present to filter out potential prompt injection attempts within the pasted threads.
Audit Metadata