manage-segments
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external data (keys) from sources such as CSV files, creating a potential surface for indirect prompt injection where malicious instructions could be embedded in data.
- Ingestion points: Ingestion occurs during the 'Add keys' and 'Replace all keys' workflows described in references/standard-membership.md.
- Boundary markers: The skill implements strict boundary markers, including mandatory reference-impact checks (usage-check.md) and production-aware approval gates (write-safety.md) mentioned in SKILL.md.
- Capability inventory: The skill utilizes tools for modifying environment state, such as harness_create, harness_update, harness_delete, and harness_execute as defined in SKILL.md.
- Sanitization: Instructions in references/standard-membership.md specify data sanitization through trimming, deduplication, and the use of redacted samples for verification to prevent raw data exposure.
Audit Metadata