TLDR — Slack ChatGPT Summarizer

TLDR is a serverless, Rust-powered Slack bot that turns a wall of unread messages into a concise, ChatGPT-generated summary delivered straight to your DM.

---

✨ Key Features

• Slash Command Workflow – Trigger summaries with /tldr in any channel.
• AI-Generated Summaries – Uses OpenAI ChatGPT to distill unread messages.
• Two-Lambda Architecture – Instant slash-command acknowledgement + async processing for snappy UX.
• Built with Safe, Async Rust – Tokio runtime, slack-morphism and openai-api-rs.

---

🏗️ High-Level Architecture

┌─────────┐    ┌────────────┐   SQS   ┌──────────────┐    ┌─────────────┐
│  Slack  │──►│ API Lambda │─▶Queue▶│ Worker Lambda │───►│ OpenAI Chat │
└─────────┘    └────────────┘         └──────┬───────┘    └─────────────┘
                                             │
                                             ▼
                                        ┌─────────┐
                                        │  User   │
                                        └─────────┘

1. API Lambda – Verifies Slack signatures and enqueues a summarisation job to SQS.
2. Worker Lambda – Fetches unread channel messages, asks ChatGPT to summarise them, and DMs the user.

---

🚀 Usage

1. Install the Slack App in your workspace (see Slack Setup below).
2. In any channel type:

/tldr

3. A DM will arrive with a neatly formatted summary of everything you missed. ✨

Advanced Parameters

You can tailor the summary by appending flags / key-value pairs after the command:

Parameter	Example	Description
count=<N>	/tldr count=50	Summarise the last N messages instead of just unread messages.
channel=<#channel>	/tldr channel=#general	Post the summary to a different channel (defaults to DM).
--visible / --public	/tldr --visible	Make the summary visible to everyone in the target channel.
custom="…"	/tldr custom="Write at an 8th-grade level"	Provide a custom prompt (max 800 chars) to influence the writing style.

Parameters can be combined:

/tldr count=100 channel=#project-updates --visible custom="Use bullet points and include action items"

---

🔧 Quick Start for Local Development

Prerequisites

• Rust (stable, Edition 2024)
• cargo-lambda ≥ 0.17 for local Lambda builds
• AWS CLI with a profile that can deploy Lambda + SQS
• Node 18+ & npm (only for the CDK infrastructure)
• A Slack workspace & OpenAI API key

Steps

# 1. Clone
$ git clone https://github.com/your-org/tldr.git && cd tldr

# 2. Configure environment
$ cp .env.example .env   # then edit the values

# 3. Build & test the Lambda crate
$ cd lambda
$ cargo test
$ cargo lambda build --release

# 4. Spin up a local Lambda for manual testing
$ cargo lambda watch   # default on :9000

Invoke the API Lambda locally with a sample payload:

$ cargo lambda invoke --data-file test/fixtures/slash_command.json

---

☁️ Deployment (AWS CDK)

The infrastructure/ folder contains an AWS CDK stack that provisions:

• API Gateway endpoint
• Two Lambda functions (API + Worker)
• SQS queue
• IAM roles & CloudWatch logs

Deploy in one command:

$ cd infrastructure
$ npm install             # first time only
$ npm run cdk deploy

After the stack is live, copy the API Gateway URL into your Slack slash-command configuration.

---

🔐 Configuration

Environment variables (set in Lambda or an .env file for local runs):

Variable	Purpose
SLACK_BOT_TOKEN	Bot OAuth token (starts with xoxb-…)
SLACK_SIGNING_SECRET	Verifies Slack requests
OPENAI_API_KEY	Access token for ChatGPT
PROCESSING_QUEUE_URL	URL of the SQS queue

---

🗂️ Project Layout

├─ lambda/          # Rust crate with both Lambda handlers
│   ├─ src/
│   │   ├─ api.rs
│   │   ├─ worker.rs
│   │   └─ bot.rs  # SlackBot implementation (shared)
│   └─ Cargo.toml
├─ infrastructure/  # AWS CDK stack (TypeScript)
├─ tests/           # Integration & fixture payloads
└─ README.md

---

🤝 Contributing

1. Make sure cargo check and cargo clippy -- -D warnings pass.
2. Add unit tests in #[cfg(test)] modules and doc-tests in public APIs.
3. Open a PR – GitHub Actions will run the full test & lint suite.

---

📄 License

MIT © 2025 TLDR Contributors