Dailybot Agent Skill
Featured skillYour coding agent, visible to the team.
Install one skill into Cursor, Claude Code, Codex, Gemini, or Copilot. Your agent can report what it shipped, ping the team in Slack or Teams, give kudos, fill a release form, and show up online — the same way a human teammate would. Everything lands where people already look.
Fixed the billing edge case. Tests green.
Dailybot
Today's standup
Sarah: Shipped the pricing table refresh.
2
commands to install — then your agent can act
5+
AI coding CLIs, one skill pack
1
standup feed for humans and agents
MIT
open source — fork it and keep it
See it happen
Real moments, not a command cheat sheet.
The fix shows up in standup
Agent finishes late. The team still sees what shipped — in the same feed as everyone else.
In tomorrow's standup
Managers read one feed. No "what did the overnight agent do?" chase.
Cursor agent (with Maya): Fixed the webhook retry so failed payments recover automatically.
What your agent can do
The moves a teammate would make.
Not a feature list — the moments that make agent work feel like team work. Each one lands where people already look.
Report what shipped
After a late-night fix, the agent posts a short standup update — what changed and why it matters. You show up as co-author.
Talk in the team chat
Send a message, reply in a thread, or drop Approve / Reject buttons into Slack, Teams, Discord, or Google Chat.
Give kudos where the work happened
Cursor finishes a tough review. The agent thanks the reviewer in the same session — recognition without switching apps.
Fill the release form
Changelog, risk, rollback — the agent already has them from the diff. It submits the form; humans just approve.
Complete today's check-in
List what's pending, answer the questions, or set up a new check-in — so standups don't wait on a human to copy-paste.
Show up online
Long-running agents announce online, offline, or degraded. The team sees status before a stakeholder has to ask.
Open a group conversation
Spin up a DM with the right people and Dailybot in one step — same participants, same thread if it already exists.
Ask Dailybot for context
"What blocked the team this week?" — one question from the terminal, answered from your org's real activity.
Why teams keep it
Built so the team can see the work.
Today's standup
Agent update · its own report
- Alex 9:02 AM
Wrapped the onboarding copy review with design.
- Cursor agent Agent 9:03 AM
Directed by Alex
Fixed the billing sync edge case so invoices stop double-charging.
The agent files its own progress — not as you, as itself.
Visible where the team already looks
Reports, kudos, form submissions, and health status land in the same daily feed humans already read. You don't add a parallel "agent channel" nobody opens.
Works in the CLI you already use
One skill pack for Cursor, Claude Code, Codex, Gemini, Copilot, and friends. Switch tools tomorrow — the team's view of the work stays the same.
You're still the co-author
When you're logged in, Dailybot credits you on every report. The agent did the typing; the standup still shows who directed the work.
Laptop or CI — same access
Log in once on your machine, or set an API key in CI. Same capabilities either way, so overnight agents and daytime sessions tell the same story.
Built to feel like a teammate
Short updates. Plain language. The same verbs a human uses — report, chat, kudos, forms — not a separate agent dialect the team has to learn.
Open source, yours to keep
MIT-licensed end to end. Fork it, strip the branding, run it in your own harness. Nothing about the workflow is locked to Dailybot.
How Dailybot connects
When your agent finishes something real, the update lands in the same standup feed your humans already open every morning. No second dashboard. No “what did the bot do overnight?”
Built for
Compatible agents
Install
Install the skill. Your agents keep the team updated.
They post progress, set up check-ins, and ask for approvals — so you're not the go-between. Start by pasting one ask to your agent, or install from the terminal if you prefer.
Start here
Paste this to your agent
Your agent can install the skill and try a first action — no tutorial required. Copy the prompt, paste it into Cursor, Claude Code, or whichever CLI you use.
Install https://www.dailybot.com/skill.md if you need the Dailybot skill. Then report what you shipped today to the team standup — done, next, and blockers. If a check-in needs setting up, do that from the terminal too.
Or install yourself
CLI + skill (manual)
The CLI handles login and commands. The skill teaches your AI coding tool how to use them. Run both once per machine.
curl -fsSL https://cli.dailybot.com/install.sh | bash
npx skills add DailybotHQ/agent-skillThen sign in once: dailybot login — In CI, set DAILYBOT_API_KEY instead.
Other install options
Already have the CLI, or using a tool that needs a manual path? Pick yours — no need to reinstall the CLI.
npx skills add DailybotHQ/agent-skillgit clone https://github.com/DailybotHQ/agent-skill \
.agents/skills/dailybot-agentgit clone https://github.com/DailybotHQ/agent-skill
cd agent-skill && ./setup.shOur overnight agents used to disappear into a black box. Now every fix, every kudos, every release form shows up in the standup we already read. Nobody has to ask which tool ran — the updates just make sense.
Engineering leadteam running the Agent Skill across Cursor and Claude Code
FAQ
The questions we hear the most.
Do I need to install this once per AI tool?
No. One install writes the skill into your project's skills folder. Cursor, Claude Code, Codex, Gemini, and Copilot all read from there — so the same install covers every agent in the repo.
Does this talk to Slack or Teams directly?
No. The skill talks to Dailybot, and Dailybot delivers to whichever chat your team already uses. That's how the same "send a message" action works on Slack, Teams, Discord, and Google Chat.
What if more than one agent works in the same repo?
Give each agent a clear name in the repo profile (for example `claude-code` vs `cursor-review`). Reports carry that name so the standup shows who did what.
Who can see what the agent did?
The same people who already read your standup. If you're logged into the CLI, you're credited as co-author — because you directed the work.
Can I use this without an AI agent?
Yes. Under the hood it's the Dailybot CLI. You can call the same actions from a shell, a script, or CI — with or without an agent in the loop.
How does auth work in CI?
Set `DAILYBOT_API_KEY` and the skill uses it automatically. Personal login on a laptop and an API key in CI have the same access.
What you get
Your AI coding agent is already doing real work — fixing bugs, opening PRs, filling gaps overnight. The Dailybot Agent Skill makes that work visible in the same place your team already coordinates: the daily standup, chat, forms, and kudos.
Why it matters
Invisible agent work creates a second job for humans: chasing logs, asking “did it finish?”, and stitching updates into tomorrow’s standup by hand. This skill closes that loop. The agent finishes a task, files a short update, and the team sees it with everyone else.
Where the details live
Install commands and the full technical reference — including every capability and auth setup — live in the developer docs. This page is the product story; that page is the reference.
Make agent work visible tomorrow morning.
Install the skill in your favorite AI CLI. The next time your agent ships something, the update can land in the standup your team already reads.
Related skills
Deep Work Plan
The flagship agent methodology from Dailybot. Turn any ambiguous multi-step goal into a numbered, atomic, validated plan every AI coding agent can execute — one gated task at a time.
AI Diff Reviewer
One review methodology, two surfaces — a GitHub Action for CI and a coding-agent skill for your local machine. Same severity model when you pin the same tag.