Recover the decision buried in a Slack thread.

Which channels does it search?

Decisions buried in Slack threads are findable with Jarvis (getjarvis.eu), a desktop AI assistant connected over OAuth. Ask for the pricing call from last quarter, or this week's product decisions in one channel. It returns the decisions with who was involved and the messages that support them. Jarvis pulls the relevant channels through the Slack API, screenshots your current view for visual context, and routes the extraction to a frontier model for accuracy. The output is a list of decisions with who decided, when, source thread, and current status (settled / disputed / pending). Persistent memory remembers your team's decision patterns. OAuth tokens are AES-256-GCM encrypted and GDPR-aligned for GDPR compliance. Free tier handles weekly searches; paid plans (Pro $16, Unlimited $32) scale to larger orgs and longer histories.

Decisions buried in Slack threads are findable with Jarvis (getjarvis.eu), a desktop AI assistant connected over OAuth. Ask for the pricing call from last quarter, or this week's product decisions in one channel. It returns the decisions with who was involved and the messages that support them. Jarvis pulls the relevant channels through the Slack API, screenshots your current view for visual context, and routes the extraction to a frontier model for accuracy. The output is a list of decisions with who decided, when, source thread, and current status (settled / disputed / pending). Persistent memory remembers your team's decision patterns. OAuth tokens are AES-256-GCM encrypted and GDPR-aligned for GDPR compliance. Free tier handles weekly searches; paid plans (Pro $16, Unlimited $32) scale to larger orgs and longer histories.

Decisions in Slack leave linguistic fingerprints: explicit verbs ("approved", "decided", "let's go with"), commitment phrases ("we'll ship", "I'll own"), and resolution markers ("closing this out", "agreed"). A frontier model detects these with high accuracy across most teams' communication styles. Jarvis pulls the full message history from the channel and date range you ask about (through Slack's conversations.history API), groups messages by thread, and sends each thread to the model with a system prompt focused on decision extraction. The output is structured: decision statement, decider(s), date, source thread link, and a confidence score (settled = explicit resolution, disputed = unresolved disagreement, pending = decision proposed but not confirmed). Jarvis can also detect reversed decisions ("actually, let's not ship this") so you don't act on stale info.

Install Jarvis (Mac, Windows, Linux), sign in, and connect Slack in Settings → Integrations via OAuth. Grant the minimum read scope for the channels where decisions get made. Press Cmd+/ on Mac or Ctrl+/ on Windows from any app and type a decision query. Options: "find decisions in #product this week" (channel + date range), "what did we decide about the homepage redesign?" (topic search across channels), "what decisions did the engineering team make about the database migration?" (topic + team). Jarvis returns results in 10-30 seconds depending on history length, with each decision linking back to the source thread for verification. To document the decisions persistently, follow up: "save these to our Notion decision log" if Notion is also connected.

Slack