Venice.Ai Router Openclaw
PassAudited by VirusTotal on May 12, 2026.
Overview
Type: OpenClaw Skill Name: venice-router Version: 1.5.0 The OpenClaw AgentSkills skill bundle for 'venice-router' is classified as benign. The `venice-router.py` script, which is the core of the skill, performs its stated purpose of routing LLM prompts to the Venice.ai API. It securely handles the `VENICE_API_KEY` from environment variables, accesses the network only to `https://api.venice.ai`, and manages local cost tracking data within the user's home directory (`~/.venice-router/costs/`). The `install.sh` script copies local files and does not fetch external payloads. Crucially, the Python script explicitly states 'stdlib only, no pip dependencies', significantly reducing supply chain risks. There is no evidence of data exfiltration, malicious execution, persistence mechanisms, or prompt injection against the OpenClaw agent itself. All file and network access is aligned with the skill's described functionality.
Findings (0)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Anyone using this skill with your API key can consume Venice.ai quota or generate billable API usage.
The skill requires a Venice.ai API credential, which is expected for the integration but grants account/API usage authority and may incur costs.
requires:\n bins: ["python3"]\n env: ["VENICE_API_KEY"]\n primaryEnv: VENICE_API_KEY
Use a dedicated API key if possible, protect the key in environment/config files, and set Venice budget or tier limits before broad agent use.
Sensitive prompts may be processed outside Venice-hosted private models unless private-only routing is enabled.
The artifacts disclose that some requests may cross from Venice to external model providers, even though identity is stripped.
**Anonymized**: Request is proxied to external provider (OpenAI, Anthropic, Google, xAI) with user identity stripped
Use `--private-only` or `VENICE_PRIVATE_ONLY=true` for confidential data, and review the selected model privacy label before sending sensitive prompts.
A user could overestimate the default privacy posture if they do not notice that private-only routing is optional.
The privacy claim is qualified to private models; because the router also supports anonymized external-provider models, users should not assume every automatic route has zero retention.
Venice offers **true privacy** with zero data retention on private models. Your conversations stay yours.
Treat privacy as setting- and model-dependent; enable private-only mode for sensitive work and avoid relying on broad marketing language alone.