sales-fyxer
Pass
Audited by Gen Agent Trust Hub on Jul 13, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill implements an indirect prompt injection attack surface via its self-improving memory mechanism.
- Ingestion points: The agent is instructed to read
references/learnings.mdat the start of each session (Step 1) to retrieve accumulated knowledge. - Boundary markers: The skill lacks delimiters or protective instructions to distinguish between the stored data and potential executable instructions within the file.
- Capability inventory: The skill has the ability to append user-derived information to
references/learnings.md(Step 4 and Gotchas) and can trigger other specialized skills via routing instructions. - Sanitization: No sanitization or validation is performed on the content written to or read from the learnings file.
- [SAFE]: The core instructional content is focused on troubleshooting and platform documentation, with no detected direct prompt injection or obfuscation.
- [SAFE]: Installation commands for related tools point to the vendor's own repository and are provided as manual user instructions rather than automated executions.
Audit Metadata