churn-autopsy
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from multiple external sources, creating a vulnerability to indirect prompt injection.
- Ingestion points: According to
references/inputs.md, the skill ingests client history, engagement data, support tickets, and exit feedback. - Boundary markers: There are no explicit instructions or delimiters mentioned in
references/standards.mdorSKILL.mdto prevent the agent from executing commands that might be found within support tickets or feedback strings. - Capability inventory: The skill is configured with
Bash,Write,Edit,WebFetch, andWebSearchtools as indicated in theSKILL.mdfrontmatter. - Sanitization: No sanitization or content validation procedures are described for the data before it enters the analysis phases defined in
references/analysis-framework.md.
Audit Metadata