Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
console.agent() is a drop-in API for running AI agent workflows anywhere in JavaScript, TypeScript, or Python code. It uses Google Gemini models and supports fire-and-forget or blocking execution. Built-in personas auto-route tasks to security, debugging, or architecture specialists.
Unlike Langchain or Agno, it requires zero boilerplate — drop a single call anywhere and it works with just a GEMINI_API_KEY environment variable.
When to use it
- Auditing SQL queries for injection vulnerabilities before execution
- Diagnosing why a unit test failed with structured AI output
- Validating API response payloads against a schema at runtime
- Researching known CVEs for a dependency during a security review
- Analyzing slow function performance with timing context in production
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: audit incoming search query for SQL injection before hitting the database
- 1Import @console-agent/agent and set GEMINI_API_KEY
- 2Call await console.agent.security("check for SQL injection", query)
- 3Receive AgentResult with data.severity and summary fields
- 4Check if severity is critical and reject the request early
- 5Log the audit result for monitoring
AgentResult with severity: "critical", fix: "Use parameterized queries", confidence: 0.94
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.