Exa Agent
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest and process data from the open web and third-party data providers, creating a surface for indirect prompt injection.
- Ingestion points: The skill fetches data from diverse external sources including B2B databases, financial news, website traffic metrics, and product catalogs.
- Boundary markers: The skill heavily promotes the use of
outputSchemato structure results into JSON format, which helps in delimiting external data from agent instructions. - Capability inventory: The agent utilizes the
agent_runtool and has instructions to create and execute local scripts for data processing. - Sanitization: While the skill emphasizes schema validation for structure, it does not provide specific instructions for sanitizing or escaping the content retrieved from the web before it is processed by the agent.
- [DYNAMIC_EXECUTION]: The "Batch Script Mode" section explicitly instructs the agent to "write a deterministic script" and then "run the script" to manage large-scale data enrichment tasks. This represents a dynamic execution pattern where code is generated and executed at runtime to facilitate interaction with the vendor's APIs.
- [COMMAND_EXECUTION]: The skill requires the agent to execute the scripts it generates within the local environment to perform batch operations, which is a powerful capability that increases the skill's operational footprint.
Audit Metadata