How does Lockstep know a Slack message is a decision?

It watches for threads in allowlisted channels where an owner lands on a call: durable, specific statements like 'let's do signed cookies, 24h'. Candidates land in a review queue as drafts; nothing becomes binding until a human confirms it.

Does Lockstep read all of our Slack?

No. It reads exactly the channels you allowlist; nothing is discovered or crawled. Tokens, keys, and credential-shaped strings are redacted at ingestion, and the only thing Lockstep ever writes back is a labeled DM.

Can we confirm decisions without leaving Slack?

Yes. When a decision is captured, its owner gets a Slack message with Confirm, Edit, and Reject buttons; one click makes it binding and briefs everyone working the affected code.

Lockstep for Slack

Lockstep watches allowlisted Slack channels and turns the decisions buried in threads into a binding, searchable record, then briefs every engineer and every AI agent before they touch the code.

What Lockstep captures from Slack

“Let’s cap free-tier exports at 100 rows” in #product
an eng lead resolving an API-shape debate in a thread
a scope cut agreed in a standup channel

How the loop runs

Lockstep captures the decision from Slack, routes it to its owner for one-click confirmation that makes it binding, then briefs every engineer and every AI coding agent before they touch the code, and flags any pull request that drifts from it. See the full pipeline on how it works.

How to set up Lockstep for Slack

In the Lockstep dashboard, open Connections, pick Slack under Connect a tool, and create the connection. Tools are connected once for your whole org.
Allowlist the channels Lockstep should read, by channel ID (for example C0123456789). Only allowlisted channels are read.
Keep talking the way you already do. When a thread in an allowlisted channel resolves on a call, Lockstep drafts it as a candidate decision in the review queue.
The decision's owner gets a Slack message with Confirm, Edit, and Reject buttons. One click makes it binding.
Engineers and AI agents working the affected code get the decision in their next briefing, and the PR check flags code that conflicts with it.

Example: a decision captured from Slack

Illustrative example

In Slack

A #product thread on export limits ends with the PM writing: "OK, final call: free-tier exports cap at 100 rows, paid plans stay unlimited."

Drafted decision

Free-tier exports are capped at 100 rows; paid plans have no row cap. Owner: the PM. Source: the #product thread, quoted and linked.

Approved record

The PM clicks Confirm in Slack. The decision is binding, anchored to the thread, and the next agent session that touches exports is briefed on it before writing code.

Access & trust

Read-only, allowlisted channels only. Secrets and PII are redacted before storage.

What Lockstep reads

Messages and threads in the channels you allowlist, and nothing else

What Lockstep never does

Channels you haven't allowlisted: nothing is discovered or crawled
Editing or deleting your messages: the only thing Lockstep posts is a labeled message to the decision's owner
Storing secrets: tokens, keys, and credential-shaped strings are redacted at ingestion

Full details on the security page.

Frequently asked questions

How does Lockstep know a Slack message is a decision?: It watches for threads in allowlisted channels where an owner lands on a call: durable, specific statements like 'let's do signed cookies, 24h'. Candidates land in a review queue as drafts; nothing becomes binding until a human confirms it.
Does Lockstep read all of our Slack?: No. It reads exactly the channels you allowlist; nothing is discovered or crawled. Tokens, keys, and credential-shaped strings are redacted at ingestion, and the only thing Lockstep ever writes back is a labeled DM.
Can we confirm decisions without leaving Slack?: Yes. When a decision is captured, its owner gets a Slack message with Confirm, Edit, and Reject buttons; one click makes it binding and briefs everyone working the affected code.