Veefore docs

Tools API

The registry of what an agent is allowed to call.

Generated page. Built from the API catalogue in apps/control-plane/src/routes/developer.ts. Edit the source, not this page.

A tool carries its own destination allowlist, risk rating and dangerous flag. Nothing is callable because it exists; an agent must be explicitly authorized for each tool.

#Endpoints

Endpoint Permission Notes
GET /v1/developer/code-assist none —
POST /v1/developer/code-assist/rewrite tool:update —
GET /v1/tools tool:read —
POST /v1/tools tool:create accepts Idempotency-Key
POST /v1/tools/{toolId}/disable tool:disable —
POST /v1/tools/{toolId}/agents tool:update —

#What each endpoint does

#GET /v1/developer/code-assist

Whether Veefore-hosted AI code assist is enabled, and which models it offers.

Requires no console permission.

#POST /v1/developer/code-assist/rewrite

Propose a guard() change for tools the automatic rewrite could not change.

Requires tool:update.

Request body:

json
{
  "model": "string",
  "language": "typescript | javascript | python",
  "file": "string",
  "source": "string (secrets replaced with placeholders)",
  "tools": "{ key, line, reason }[]"
}

#GET /v1/tools

List tools.

Requires tool:read.

#POST /v1/tools

Register a tool.

Requires tool:create.

#POST /v1/tools/{toolId}/disable

Disable a tool for every agent.

Requires tool:disable.

Request body:

json
{
  "reason": "string"
}

#POST /v1/tools/{toolId}/agents

Authorize an agent for a tool.

Requires tool:update.

Identity, authority, policy and enforcement for autonomous AI agents.