byted-byteplus-infoquest-search
Pass
Audited by Gen Agent Trust Hub on Mar 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill uses environment variables (
INFOQUEST_API_KEY) for authentication, following security best practices for credential management. - [COMMAND_EXECUTION]: Provides Python CLI scripts (
scripts/search.pyandscripts/extract.py) to interface with external web services. - [EXTERNAL_DOWNLOADS]: Recommends the installation of the
requestslibrary viapipto facilitate HTTP communication. - [DATA_EXFILTRATION]: Communicates with external endpoints at
reader.infoquest.bytepluses.comandsearch.infoquest.bytepluses.com. These are legitimate vendor-owned resources associated with the skill author, ByteDance. - [PROMPT_INJECTION]: The skill acts as an ingestion point for untrusted external data (search results and webpage content), which presents a surface for indirect prompt injection.
- Ingestion points: Raw webpage content and search snippets are ingested into the agent context via the BytePlus API responses.
- Boundary markers: Not present. Data is returned in JSON or formatted text blocks without explicit isolation instructions.
- Capability inventory: The skill's scripts are restricted to making API calls and printing results; they do not possess dangerous capabilities like local file writes, subprocess execution, or arbitrary code evaluation.
- Sanitization: The scripts perform field extraction and JSON parsing but do not sanitize the text content for potential injection patterns.
Audit Metadata