algolia-core-workflow-b
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from a source dataset, which serves as an entry point for untrusted content that could influence the agent's behavior if not correctly handled.
- Ingestion points: Data enters the context via the
source-datasetargument and subsequent file-reading operations mentioned in the instructions (SKILL.md). - Boundary markers: The instructions include explicit guidelines for the agent to "Transform and validate records locally" and check for "prohibited data," which provides a level of behavioral boundary control.
- Capability inventory: The skill enables the use of
Write,Edit, andWebFetchtools (SKILL.md), representing a surface where successful injection could lead to unauthorized file modifications or network activity. - Sanitization: The skill advises the agent to perform local validation (size, IDs, required fields) but does not provide a formal sanitization or escaping mechanism for the data before it is processed (SKILL.md).
Audit Metadata