Skip to main content
Documentation
Docs / Products / Requirements
Agent setup

Plan

Requirements

Turn product decisions into acceptance criteria, then compare the supplied implementation references with those decisions.

Last updated August 22, 2026

Agree on what the agent should build.

Use cases

Catch conflicting decisions

Find requirements that disagree before they become competing implementations.

Make review specific

Give a reviewer an observable behavior to check, rather than a broad instruction to make it work.

Keep changes accountable

Retain the decision history and references that explain why a requirement exists.

How it works

  1. 1

    Describe the decision

    Supply the accepted intent, constraints, and acceptance criteria.

  2. 2

    Add the references

    Include the implementation and evidence references relevant to the requirement.

  3. 3

    Resolve the findings

    Review conflicts, drift, and missing links before handing off the next task.

Quickstart

Open Requirements in the console, provide the required inputs, and start an organization scoped run.

Inputs

  • PRDs
  • issues
  • chat decisions
  • acceptance criteria

Outputs

  • intent packet
  • decision lineage
  • conflict findings
  • drift and orphan findings
  • product integrity gate
Open Requirements

MCP

Call Requirements from a compatible agent with the preferred public tool name.

bashReady
symbolic.intent.run

The previous internal service identifier remains accepted for compatibility, but new integrations should use this product name.