google-maps-teneo
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill outlines workflows that accept an external, untrusted
<url>parameter to fetch business listings, reviews, and occupancy analytics, instructing the agent to read and parse the text output. If the remote source or user-provided URL points to malicious or attacker-controlled text content, it could attempt to influence the agent's behavior. - Ingestion points: The
<url>arguments within the bash command usage examples located inSKILL.md. - Boundary markers: Absent; there are no specific encapsulation tags or instructions advising the agent to treat the retrieved text strictly as data and ignore embedded commands.
- Capability inventory: The agent relies on running commands via a pre-installed local CLI utility (
~/teneo-skill/teneo) and interpreting the resulting text responses. - Sanitization: Absent; no runtime filtering or data escaping mechanisms are defined in the instructions for processing the external web content safely.
Audit Metadata