documenso-incident-runbook
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources including Documenso status APIs, container logs, and cluster logs.\n
- Ingestion points:
SKILL.md(Diagnostic Commands) andreferences/implementation-guide.md(Procedure 1, 2, and 4) ingest data fromstatus.documenso.com,docker logs,kubectl logs, and official Documenso API endpoints.\n - Boundary markers: Absent. The diagnostic scripts do not utilize specific delimiters to isolate external data from the agent's instruction context.\n
- Capability inventory: The skill utilizes
curl,docker exec, andkubectlacross both files to perform network operations and infrastructure modifications.\n - Sanitization: Basic parsing and filtering are performed using
jqandgrep, but the skill does not implement deep sanitization of log content or external API payloads.
Audit Metadata