Roadmap
Enforcement ships first. Here's what's next.
IntentKeep starts by locking requirements and blocking the changes that break them. These are the capabilities we're building on top of that foundation.
Three Amigos review
Polish every new requirement before it reaches a developer.
Add a requirement and run a Three Amigos review. An AI business analyst, tester, and developer pressure-test it from every angle - whether it is worth building, whether QA can even test it and which edge cases it misses, and whether it is feasible against your connected codebase - then hand back a sharper, more complete requirement that is ready to build.
Business analyst
Challenges the value and scope - whether the requirement is worth building and how it fits your priorities - so effort goes to what moves the needle, not noise.
Tester
Makes sure the requirement is genuinely testable - sensible naming, test IDs on the controls, a scenario QA can actually verify - and surfaces the edge cases it misses, so untestable asks and missing scenarios become requirements now, not bugs in production.
Developer
Weighs the requirement against your connected codebase and flags what is feasible, what is costly, and what is not possible yet.
new requirement → Three Amigos review → sharper, build-ready requirement
Build from intent
Your AI builds from locked intent, not guesswork.
Connect IntentKeep to Cursor, Claude, or Codex over MCP. Your locked requirements travel into the tools where code is written, so the AI builds against what you agreed and stays on track.
Locked intent, not loose prompts
Your approved requirements are the source of truth. The AI builds against what you agreed, not a half-remembered chat.
Connects over MCP
IntentKeep plugs into Cursor, Claude, and Codex through MCP, so your intent travels into the tools where code is actually written.
Keeps your AI on track
Every build references the locked requirements, so your agent stays inside the lines you drew instead of drifting away from them.
locked requirements → MCP → your coding agent
Software Intent Audit
Find out where agreed behaviour quietly stopped holding
One application. Two weeks. You leave with the behaviours the software currently guarantees, the places they contradict each other, and pull request checks on the ones you lock.
- Software Intent Audit · £1,750
- Behaviour Snapshot and Guard · £950