algo-hr-compensation
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection due to its core function of processing external, untrusted datasets.
- Ingestion points: Untrusted data enters the agent context in
SKILL.md(Phase 1), where the agent is instructed to collect internal compensation data and market survey data. - Boundary markers: The skill body lacks explicit delimiters or instructions to ignore embedded commands within the ingested HR or survey data.
- Capability inventory: The skill provides Python reference code in
references/pay-equity.mdinvolving data processing libraries, which suggests the agent will perform computations on the ingested data. - Sanitization: There are no specific instructions or algorithmic steps provided to sanitize input data or validate it against a schema to prevent instructions embedded in data fields from influencing the agent's behavior.
Audit Metadata