Zest
Zest is a VS Code extension that tracks your AI-assisted coding sessions and generates automated standups. It reads your AI agent conversations in Cursor and GitHub Copilot to provide insights into your development workflow.
Features
⭐️ AI Metrics
Track and analyze your AI-assisted coding sessions with detailed metrics:
- AI Agent Conversations: Captures conversations with AI assistants (Cursor, GitHub Copilot) to understand your workflow
- Usage: Models, token counts and tool use reported by the assistant
🤖 Automated Standups
Forget about writing standups — Zest generates them for you:
- AI-powered analysis of your coding sessions
- Daily summaries of your development activity
- Ready-to-share standup reports
How It Works
💻 Your IDE Activity → 🍋 Zest Collects → 🔒 Secure Backend → 🧠 AI Analysis
What we collect:
- ✓ AI agent conversations: messages, titles, and the metadata the assistant stores with them
(model, token counts, tool use)
- ✓ Which roles are uploaded is set by your workspace's data controls
What we DON'T collect:
- ✗ Sensitive data (API keys, passwords): always detected and redacted before upload
- ✗ Browsing history, file contents you did not paste into a chat, or editor events
📡 Sync Status
The Zest sidebar shows whether syncing is actually working, as its first item — so
a stalled sync is visible rather than silent:
- Sync: OK with the time of the last successful sync
- Sync paused with the reason (no workspace yet, chat tracking unavailable,
turned off in settings, message collection disabled, server unreachable)
- Sync error with the reason (chat extraction failed, data policy unavailable,
server rejected the upload)
- Sign-in required, which signs you in when clicked
Privacy Controls
Your data, your control:
- Sensitive data redaction: API keys, passwords, and secrets are detected and redacted from
every message, title and metadata field before upload. It cannot be turned off; you choose the
replacement text (a pattern mask, a hash, or both)
- Detections list: Zest: Show Sensitive Data Detections lists what was redacted (the
pattern and where, never the text)
- Your settings, not the repo's: privacy settings are read from your user settings; a
workspace's
.vscode/settings.json can only make them stricter
Requirements
- VS Code 1.95.0 or higher (or Cursor)
Getting Started
- Sign in to connect your Zest account: the browser opens with a code filled in, you approve
it, and the extension signs in (no redirect back to the editor needed)
- Start coding with your AI assistant
- Generate your standup when you're ready
Access the welcome screen anytime: Command Palette → Zest: Welcome
Extension Commands
| Command |
Description |
Zest: Welcome |
Open the welcome screen |
Zest: Getting Started |
View getting started guide |
Zest: Sign In |
Sign in to your Zest account |
Zest: Sign Up |
Create a new Zest account |
Zest: Sign Out |
Sign out of your account |
Zest: Show Daily Summaries |
Open your standups in the Zest web app |
Zest: Open Privacy Settings |
Configure privacy options |
Zest: Show Sensitive Data Detections |
List what was redacted before upload |
Zest: Configure Detection Sensitivity |
Turn aggressive detection (emails, IPs, etc.) on or off |
Extension Settings
| Setting |
Default |
Description |
zest.remotePersistence |
true |
Upload captured sessions to Zest (user setting; a workspace can only turn it off) |
zest.privacy.approach |
"detection" |
Replacement text: detection (pattern mask), encryption (hash), or hybrid |
zest.privacy.aggressiveMode |
false |
Also redact emails, IPs, etc. |
Development
Builds talk to https://meetzest.com; ZEST_BUILD_API_URL bakes another base. To run against a
local API (wrangler dev), set ZEST_API_URL=http://localhost:8787 in the Extension
Development Host's environment (e.g. env in launch.json); it is read at runtime and never
baked into a build.
Start tracking your AI-assisted coding journey! 🍋
| |