Capture Slack and email discussions so a team can later ask why a decision was made.
Connect it to Cursor via MCP to ask about past decisions while coding.
Ingest Jira, GitHub, and Linear webhooks to build a searchable decision history.
| sunradiance/whyline | 0xustaz/streamgate | a-bissell/unleash-lite | |
|---|---|---|---|
| Stars | 1 | 1 | 1 |
| Language | Python | Python | Python |
| Setup difficulty | moderate | hard | hard |
| Complexity | 3/5 | 4/5 | 4/5 |
| Audience | developer | developer | researcher |
Figures from each repo's GitHub metadata at analysis time.
Requires an LLM API key and local setup, production team use needs a WSGI server like gunicorn.
Whyline is a self-hosted tool that helps a company keep track of why past decisions were made, so that months or years later someone can ask why a rule or policy exists and get a real answer with sources, instead of everyone forgetting and re-investigating from scratch. The README describes the common problem of a team repeatedly asking things like why a certain billing rule exists or why a vendor was rejected, with the reasoning lost because the person who knew has left or the conversation is buried in an old Slack thread. The tool stores decisions in a shared SQLite database, a lightweight file-based database, so the information is available to the whole team rather than trapped in one person's browser. It uses a search technique called BM25 to find the most relevant past decisions for a given question, which is designed to keep working even as the collection of decisions grows into the hundreds. Every answer it gives links back to where the information came from, such as a Slack message, an email, a document, or a Jira ticket, so the source can be checked. Decisions can be captured automatically from several places: Slack messages and threads, an email alias, meeting transcripts, documents in Notion, Confluence, or Google Docs, and webhooks from GitHub, Microsoft Teams, Linear, Jira, and Salesforce. It also works as an MCP server, a standard that lets AI coding assistants like Cursor connect to outside tools, so a developer can ask questions about company decisions directly inside their editor. Setup involves cloning the repository, copying an example environment file, adding an API key for a language model service, and running a setup and start command, after which the tool runs locally on your own machine at a local web address. The project also supports multiple user roles for team use, such as viewer, member, admin, and owner, each with different permissions for creating and managing decisions. It is released under the MIT license with no required payment, though it accepts optional voluntary support.
A self-hosted tool that captures why past company decisions were made from Slack, email, and other sources, so anyone can later ask why and get sourced answers.
Mainly Python. The stack also includes Python, SQLite, Flask.
Licensed under the MIT license, so it can be used, modified, and shared freely including for commercial purposes.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Don't trust strangers blindly. Verify against the repo.