Your AI agents know the code.
They don't know the company.
The decisions behind your code live in meetings, chat and tickets. Align turns them into one record of what still stands, what conflicts, and why - checked by your team and its agents before they build, and the input your ADRs draw from.
- 65%
- of developers using AI to refactor say it misses relevant context1
- 46to95%
- decision compliance in a published benchmark, once the agent was given recorded decisions2
Start with the free CLI
Open source under MIT, no account, all integrations included. It builds your own decision graph and serves it to any MCP agent.
curl -fsSL https://align.tech/install.sh | sh Want it for your whole team? Tell us where to send the hosted version.
You can run the whole thing yourself
Plain containers on your own infrastructure, against your own Postgres and Redis. The graph lives in your database and Align Tech Ltd has no access to it. With no cloud API keys set it makes no external LLM calls.
From a meeting decision to every agent that builds next - before anyone ships against it.
One decision, followed all the way from the meeting it was made in to the pull request that ships against it.
you What's the story on how we handle customer data in logs? How did we end up here?
align Three decisions, two tools, one supersession. The security review banned customer PII from application logs (Teams, 12 Mar). The same meeting put redaction in the log shipper - superseded. It moved to redacting at source three weeks later (Jira, 2 Apr), which is what stands today.
- 1 Teams meeting
"Security review: no customer PII goes into application logs. That's every service, no exceptions - redact it before it's written."
@align capture this
- 2 Decision capturedNo customer PII in application logsActive Security review
Redact customer PII before anything is written to an application log. Every service, no exceptions.
AI extracts the decision and its rationale. You confirm it - no template, no doc.
- 3 Conflict caught across tools
Slack thread · weeks later
The payments squad agrees to log full request bodies, unaware of the org-wide ban.
Conflict a team-level choice contradicts the org decision
Cross-tool, cross-team detection - the contradiction is caught as the decision enters the graph, before either side ships against it.
- 4 Linked into the graph
- 5 Every agent builds on it
agent Adding request-body logging to the checkout service to trace the failure...
align Hold - the security review banned customer PII from application logs, and a raw request body carries it. Redact at source before you log.
- 6 And again at the PR
Align Decision Check · pull request #482
Logs the raw request body, against the org-wide ban. Every other check on the PR is green.
Blocked a required status the merge cannot pass
Decision capturedAnd it comes back round
- Where decisions get made
- The decision graph
- People and agents
- The pull request
A merged pull request isn't a new decision - it's evidence the decision held, so it goes back into the graph as the record of what shipped.
Up and running in minutes: connect your tools over OAuth, then @align in chat, /align in comments, or react with
an emoji - your own, per channel or org-wide. No new tool to open, no training
- just the trigger you were already using.
Your decision graph, kept current across every tool.
Every decision your org made, from the big architectural calls to the day-to-day ones settled in chat, linked across tools and classified by how far each one reaches.
Decisions linked across tools, sized by architectural reach - system-wide, one component, or a local detail
One scan surfaces the decisions already buried in your tools.
Align reads back through your connected tools and reconstructs what your team already decided, then shows what still stands, what got superseded, and what conflicts.
Dogfooded on Align's own engineering workspace - not a synthetic benchmark
Our own GitHub org, 22 August 2026. Candidates are unreviewed - the adjudicated number follows once we can state its method.
Start on your own, or see it on your data.
Try it yourself
For one engineer, on their own history.
Point it at a repo and read the graph it builds.
curl -fsSL https://align.tech/install.sh | sh See it on your data
For a team, across all your tools.
We connect your tools, run the historical scan, and walk you through what it found.
Book a demo Self-hosting? Talk to usThe intent layer your agents are missing.
Align captures decisions where they happen, links them across every tool, and exposes them through an MCP server. Agents get warned before they build, engineers search it before they debate, and every pull request gets checked back against the same graph.
Answered before it writes
Your agent asks over MCP and gets the current answer with its source, from any repo, tool or team.
Warned before you build
Align flags the decision an agent is about to break, surfaced unasked, before deployment.
Checked again at the PR
Align compares the pull request against the decisions it touches. Make the Align Decision Check a required status and a conflict blocks the merge.
Connects where decisions happen
Your decision graph is yours.
We won't pretend to reverse-engineer a decision from code where nobody recorded the why. Nobody can. What we promise is that from the day you connect Align, the reasoning stops disappearing.
Open-source core
The CLI and connector SDK are MIT-licensed. Read the code, run it yourself, no account required.
We can't read your graph
Run it yourself and Align Tech Ltd has no access to your decisions. Hosted, we store them in eu-west-2.
Your messages aren't the product
Align only processes conversations you invite it into, extracts the decision, and discards the raw text. We store decisions, not your messages.
Simple plans, all integrations included
Start free with the CLI. Partner with us to get the shared organizational memory across your whole team - all integrations included on every plan.
Personal use. Import decisions from every tool you already use - then expose your local graph as a provider-agnostic MCP server any AI agent can read.
- All integrations included - Slack, GitHub, Jira, Linear & more
- Personal decision graph
- Provider-agnostic MCP server (Claude, Cursor, any agent)
- Open source (MIT)
- No account required
Full product, all integrations, and direct founder support for teams helping shape Align. The shared decision graph across your whole org - captured from every tool your teams already use.
- Everything in CLI, for your whole team
- Shared, org-wide decision graph
- Conflict detection & alerts
- MCP server for every agent
- Direct founder support & roadmap input
For teams that need self-hosting, SSO, or compliance. Run Align in your own infrastructure, where Align Tech Ltd has no access to your decision graph.
- Everything in Pro
- Self-hosted via Helm, air-gap supported
- OIDC SSO + SCIM provisioning
- SOC 2 & ISO 27001 roadmap
- Dedicated support & SLA
No integration limits on any plan. Licensing for Enterprise & Self-Hosted is based on team size.
Or start on your own, free:
curl -fsSL https://align.tech/install.sh | sh Go Deeper
The technical architecture behind the decision graph. How Align captures decisions across the SDLC, structures them with AI, surfaces conflicts and superseded decisions across tools, and gives your agents the current answer with its source.
Read the Technical Whitepaper →