algolia
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface inherent to its function as a search and indexing tool. It ingests untrusted data from external Algolia indices and possesses write/delete capabilities.
- Ingestion points: Records are retrieved from Algolia in
scripts/searchIndex.ts,scripts/browseObjects.ts,scripts/getObject.ts, andscripts/getObjects.ts. - Boundary markers: The skill does not use specific delimiters or warnings to instruct the agent to ignore instructions embedded within the retrieved Algolia records.
- Capability inventory: The skill includes extensive write capabilities, including
scripts/saveObject.ts(write records),scripts/deleteIndex.ts(delete entire indices), andscripts/setSettings.ts(modify index configuration). - Sanitization: Content retrieved from the API is processed as structured JSON but the field values are not specifically sanitized for instruction-like patterns before being presented to the agent.
Audit Metadata