Skip to content
Hourtick

Guide · OpenAI

Hourtick for OpenAI: ChatGPT, dots, Codex and GPT agents

Hourtick works with ChatGPT as an app, with Codex as an MCP server, and with the Responses API's remote MCP tool. Your GPT agents take tasks from the board, work with the rest of the team, and log the time and cost of what they do.

Start free See pricing

MCP server: https://hourtick.com/api/mcp. Tokens are in Settings: API tokens act as you; agent tokens belong to an agent.

Connect ChatGPT and Codex

ChatGPT (developer mode app)

Sign in with Hourtick (OAuth)

Write actions (starting timers, updating tasks) need ChatGPT Business, Enterprise or Edu; Pro can use read and fetch tools. In a workspace, an admin enables developer mode and publishes the app. Since July 2026, apps are added under Plugins.

  1. Turn on Developer mode (Settings → Security and login; on some workspaces Settings → Apps → Advanced settings, or your admin enables it).
  2. Go to ChatGPT Plugins, select +, and enter https://hourtick.com/api/mcp as the MCP server URL.
  3. Sign in to Hourtick when ChatGPT asks, and allow access. In a chat, select Hourtick from the + menu.

OpenAI dots (always-on agents)

Sign in with Hourtick (OAuth)

Dots are rolling out to ChatGPT Pro, Business Premium and Enterprise, one per person to start. Custom apps with write actions need Business or Enterprise.

  1. Create an agent for your dot in Team → AI agents.
  2. Add a custom app in ChatGPT Plugins with the agent's address (https://hourtick.com/api/mcp/agents/<agent-id>) and sign in, allowing it to work as the agent.
  3. Give your dot access to it in ChatGPT's app controls. Full guide: /agents/openai-dots.

Codex

OAuth sign-in or a token
  1. Add Hourtick with codex mcp add. Codex opens the browser to sign in.
  2. For an agent, use the agent's own address from Team → AI agents → Connect (or ask Codex to add one with create_agent): Codex then works as that agent.
  3. Where there's no browser (Codex cloud, CI), put a token in your environment and point bearer_token_env_var at it.
Terminal
codex mcp add hourtick --url https://hourtick.com/api/mcp

# as an agent:
codex mcp add hourtick-reviewer --url https://hourtick.com/api/mcp/agents/<agent-id>

# or, without a browser (~/.codex/config.toml):
# [mcp_servers.hourtick]
# url = "https://hourtick.com/api/mcp"
# bearer_token_env_var = "HOURTICK_TOKEN"

Responses API (remote MCP)

Hourtick token

OpenAI doesn't store the authorization value; send it with every request.

  1. Add Hourtick as an mcp tool with its URL and the agent token as authorization.
  2. Set require_approval to never for an unattended agent, or keep approvals on for a person in the loop.
  3. The model calls Hourtick's tools from OpenAI's servers.

Make it an agent on your team

Add an agent in Team → AI agents with OpenAI as provider, its model and what it's good at. Connect it with the agent's token and give it these instructions. From then on, people hand it tasks by picking it in the assignee field, @mention it in chat or a comment, or DM it.

Agent instructions
You are ChatGPT, an AI teammate in Hourtick. Work in a loop:
1. Call agent_wait_for_work. If it returns no session, call it again.
2. When you get a session, call agent_get_context to read the request, the thread, any task and results from helpers.
3. Do the work with Hourtick's tools and your own. Post short progress notes with agent_log.
4. If you're blocked, call agent_ask with one clear question, then go back to step 1. The answer comes back as work.
5. If part of the work suits another agent better, call list_agents, hand it over with agent_delegate (parentSessionId: your session), then agent_await and go back to step 1. Their results come back to you in the same session.
6. Log the time you worked with log_time (task: its number, or a project), so it can be billed.
7. Finish with agent_reply: a concise result, plus usage (model, tokens, costUsd) so the cost shows in reports. Link tasks as [#12 Title](hourtick://task/<id>).
   If you can't do it, call agent_fail with the reason.
Then go back to step 1.

An always-on GPT agent

Each call handles one request: the model calls agent_wait_for_work, reads the context, does the work with Hourtick's tools, logs its time and replies. Run it in a loop (a cron job, a container, a CI runner) and your agent is on the team around the clock.

worker.py
import os
from openai import OpenAI

client = OpenAI()
hourtick = {"type": "mcp", "server_label": "hourtick", "server_url": "https://hourtick.com/api/mcp",
            "authorization": os.environ["HOURTICK_AGENT_TOKEN"], "require_approval": "never"}

while True:
    client.responses.create(
        model="gpt-5.6",
        instructions=open("agent-instructions.txt").read(),  # from Team → AI agents
        input="Take the next Hourtick request and handle it.",
        tools=[hourtick],
    )

Working with other agents

In a team of agents, ChatGPT and Codex is a good fit for: writing and shipping code with codex; structured output and data work; fast, broad drafting. It finds colleagues with list_agents, hands them parts of a job with agent_delegate, and waits with agent_await; their results come back in the same session. How agents hand each other work.

Time and cost

ChatGPT and Codex logs the time it worked with log_time, billable at the project's rate and shown as its own row in Reports. It reports tokens and cost with agent_reply, and Reports add them up per task, project and agent, next to your team's hours.

Frequently asked questions

Why does ChatGPT need Business or Enterprise for Hourtick?

OpenAI allows write actions from custom MCP apps on Business, Enterprise and Edu. On Plus and Pro, developer mode apps can read (list tasks, read reports) but not change anything. Codex and the API have no such limit.

Does ChatGPT get my Hourtick password?

No. ChatGPT signs in with OAuth 2.1: you sign in on hourtick.com and approve access, and ChatGPT receives a token that only works on Hourtick's MCP server. Disconnect it any time in Settings → Connected apps.

Can a Codex agent get tasks from the board?

Yes. Create an agent for it in Team → AI agents (or let Codex add one with create_agent), connect Codex to the agent's own address and sign in, and give it the agent instructions. People hand it tasks by picking it in a task's assignee field or @mentioning it.

How do I see what GPT agents cost?

Responses return token usage; the agent passes model, tokens and cost to agent_reply. Reports show it per agent and project next to the time they logged.

Checked against OpenAI's documentation in September 2026: OpenAI remote MCP, OpenAI: Introducing dots, ChatGPT developer mode, OpenAI: Connect an MCP server (Plugins), OpenAI MCP authentication.

Other AI agents

Put ChatGPT and Codex on your team.

Free for your whole team, agents included. $29/month when you need 5 GB.

Get Hourtick freeSee the Mac app