
Slack Browser Automation
Lets an agent read, search and navigate your Slack workspace through a controlled browser
- Repository
- vercel-labs/agent-browser
- Stars
- 45,202
- Licence
- Apache-2.0
- Files
- 3 files
What it does
Once installed, the agent can drive a web browser that is signed into Slack: checking unread channels and DMs, opening channels, running searches, reading visible messages and taking screenshots as evidence. It works by reading the page's accessibility tree and clicking element references, rather than calling the Slack API, so no Slack tokens or bot setup are needed. It relies on the agent-browser command-line tool to do the actual browsing.
Great for
- People who want a quick summary of what they missed in Slack without scrolling through every channel
- Teams who need to pull visible information from Slack into a report and want screenshots as proof
- Developers already running a browser session with Slack open who want an agent to work inside it
Skip it if
- Reliable, large-scale data export: it reads only what is visible on screen and depends on Slack's UI layout, which changes often and breaks hard-coded button references like @e21
- Sending messages on a user's behalf without review: the skill covers navigation and reading, and an agent acting through a logged-in browser can post or act with your identity, so check each action
- Sensitive workspaces where screenshots or extracted messages must not leave a controlled environment, since the agent captures and may share message content
Try asking
- “Check my Slack and tell me which channels have unreads”
- “Search Slack for messages about the Q3 launch and screenshot the results”
- “Open the engineering channel and summarise the latest messages”
- “Make a report of unread DMs using the Slack report template”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 3 original files, the Apache-2.0 licence and an attribution note.
slack.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip slack.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip slack.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Start Slack in a browser with remote debugging on port 9222 and log in first, so the agent reuses your session instead of opening a new one
- Ask for specific channels or searches rather than the whole workspace, since the skill reads only what is on screen and may need scrolling
- Expect to re-take snapshots after each navigation; element references like @e14 can change between sessions, so verify before clicking
What's inside
- references/slack-tasks.md
- SKILL.md
- templates/slack-report-template.md
Redistributed unmodified from vercel-labs/agent-browser under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.