AI agents call tools and retrieve context with no runtime governance. VisIQ fixes that in two lines of code.Documentation Index
Fetch the complete documentation index at: https://docs.visiqlabs.com/llms.txt
Use this file to discover all available pages before exploring further.
executor.invoke() calls work unchanged — but every tool call and every retrieved document is now governed.
What activates
Action Governance
Controls what agents can do. Every tool call is evaluated against your rules before it executes — permit, deny, or route to a human for approval.
Context Governance
Controls what agents can see. Every retrieved document is filtered against your policy before it reaches the agent’s context window.
Audit Trail
Proves what agents did. Every decision is cryptographically signed and hash-chained into an immutable ledger.
visiq() call. There is no additional code, configuration, or per-tool wrapping.
How it works
visiq() detects your framework
The harness inspects the target object. It recognizes LangChain
AgentExecutor and LangGraph CompiledGraph.Action governance installs at handleToolStart
Before any tool function executes, the proposed call (tool name + arguments) is evaluated against a locally cached rule bundle. Sub-millisecond on the hot path — no network round-trip.
Context governance instruments retrievers in place
The harness walks the
tools array, finds retrievers, and monkey-patches them. Every returned document passes through policy evaluation before the agent sees it.Install
.env
VISIQ_API_KEY, VISIQ_ENDPOINT, and VISIQ_AGENT_ID from environment variables automatically. Options passed to visiq() take precedence.
Next steps
Quickstart
Full working example with tools and a retriever — 5 minutes.
SDK Reference
Complete
visiq() API — options, framework detection, error behavior.