Decawork
Approval and access layer for AI agents: agents are registered with an identity and owner, their API and MCP traffic is routed through a gateway, and a model trained on the company's plain-English policies allows, blocks or holds each proposed action for human review, then revokes access when the task is done; booked through a demo.
What it does
- Guardrails and safety
- Agent orchestration
Features 4
- Plain-English policiesDefine what agents may do and when a person must step in.
- Decisions before executionEach proposed action is allowed, blocked or held for human review.
- Agent registryEvery agent keeps an identity, an owner and visible activity.
- Task-scoped accessAccess is reassessed as the agent works and revoked when the task is complete.
Sources
- Pricing
-
Official documentation · 19 Oct 2026
Home page: calls to action read Book a demo and Book a live walkthrough; the terms say a pilot or paid product is governed by its own written agreement. No plan or price is published.
- Description
-
Official documentation · 19 Oct 2026
Home page: "We train a model for your company to keep agent actions aligned with your interests."
- What it does
-
Official documentation · 19 Oct 2026
Home page: "We train a model for your company to keep agent actions aligned with your interests."
- Sold within
-
Sold through a demo, walkthrough and pilot; no self-serve plan or price appears on the pages read.
- Name
-
Official documentation · 19 Oct 2026
Name as published on the product page.
- URL
-
Official documentation · 19 Oct 2026
Product page, read 2026-10-19.
- Deployment
-
Official documentation · 19 Oct 2026
The pages read state no deployment model beyond the vendor-run service; saas is recorded as the vendor-operated reading.
- Interfaces
-
Official documentation · 19 Oct 2026
Security page: "Route API requests and MCP tool calls through Decawork."
Open Decawork in the directory
Something wrong here? Send a correction — quote this product id: decawork-agent-approvals.