The agent-native issue tracker

Your AI agent is ready to work. Stop managing it in a chat.

Give it an Epic. Issopen turns the plan into a shared queue, lets agents claim and complete work, pauses for your decisions and brings every result back for human review.

MCP-native · Human-reviewed · Self-hostable

⌁ Codex claimed 42 Working from the approved plan
✓ Ready for your review Commit and evidence attached
app.issopen.com Agent online
Illustrative Issopen board for The One-Prompt App, with tickets about lost AI chat context

One backlogNo more context scattered across chats.

Any MCP agentBring Codex, ChatGPT or your own client.

You stay in controlAgents execute. Humans make the call.

The chat trap

A chat is a conversation.
It is not a workflow.

Every session starts with another explanation. Decisions vanish into scrollback. Two agents can unknowingly do the same work. When the answer arrives, nobody knows what was verified.

01

Context resets

Your agent forgets the plan when the session ends. You become the copy-and-paste layer.

02

Work stays invisible

No shared queue, no ownership and no reliable answer to “what is the agent doing?”

03

Questions block everything

One missing decision stalls the whole conversation instead of only the affected ticket.

04

Review becomes archaeology

You hunt through messages, branches and screenshots just to understand what changed.

The shift is already happening

The model gap is closing. The workflow gap is not.

The teams that move fastest will not be the ones with the longest prompts. They will be the ones whose agents wake up to a clear queue, know when to ask and return work that is ready to review.

From intent to reviewed code

Give your agent a place to work.

Issopen turns autonomy into a visible loop you can trust.

  1. 01

    You set the direction

    Drop an outcome, not a giant prompt.

    Create an Epic, add the tickets you already know and answer the decisions that shape the work. The plan stays beside the execution instead of disappearing into a transcript.

  2. 02

    Your agent runs the queue

    It claims, builds, verifies and reports.

    Through MCP, the agent sees only the projects and actions you allowed. It claims one ticket, records progress, links code and keeps moving through independent work.

  3. 03

    You keep the final say

    Questions and results come back to you.

    When judgment is required, Issopen shows a warning, a focused question and the agent's recommendation. Completed work returns with the evidence you need to approve it.

Capture the moment it breaks

See it. Paste it. Turn it into work.

Product feedback loses energy every time someone says “I will write the ticket later.” The Issopen Chrome extension turns a screenshot into a structured issue while the context is still fresh.

  • Paste a screenshot or upload several images
  • Choose the project and Epic without leaving the page
  • Keep drafts local until you explicitly send them
Explore the Chrome workflow
Illustrative Issopen Chrome extension drafting an issue after an AI forgot a decision from chat

Autonomy without surrender

Let agents move fast.
Keep the boundaries explicit.

Issopen gives an agent enough access to do useful work, not a master key to your workspace.

◎

Scoped by project and capability

Decide which projects an identity can see and whether it may read, write, claim, create, review or close work. Change those permissions without rebuilding the integration.

↺

Every action leaves a trail

Claims, comments, decisions, status changes and code links stay attributed and visible.

◇

Humans own the finish line

Require human review before Done. Agents deliver the evidence; your team decides.

⌁

Built on MCP, not locked to one model

Connect Codex, ChatGPT or another compatible client. Your backlog outlives any single chat, model or agent session.

Read the agent onboarding guide

Built for the teams leaning forward

AI can write the code.
Issopen makes it part of the team.

For founders

Turn product intent into continuous execution.

Keep the roadmap, decisions and agent output in one loop you can inspect at any time.

For product teams

Move from visual feedback to assigned work instantly.

Capture what is wrong, attach the evidence and let the right human or agent pick it up.

For developers

Delegate the queue without delegating judgment.

Let agents handle bounded tasks while architecture and final review stay with your team.

Do not give your agent another blank chat

Give it a queue.
Give it context.
Let it come back with proof.

The opportunity is not using AI once. It is building a system where useful agent work can happen again tomorrow without starting over.

Questions worth asking

Before you give an agent a backlog.

Does Issopen run or host the AI model?

No. You bring the agent or MCP client you already use. Issopen coordinates work, permissions, decisions and review.

Can an agent change everything in my workspace?

Only if you explicitly grant those capabilities. Agent identities are scoped by project and action, and their access can be reduced or revoked.

What happens when the agent needs a decision?

It can add a visible question with a recommendation and bounded options. You answer inside the ticket while independent work can continue elsewhere in the Epic.

Can I use Issopen without an AI agent?

Yes. It is a complete project, Epic and ticket workflow for humans. Agent collaboration is an additional teammate, not a requirement.

Can I host Issopen myself?

Yes, for uses permitted by the PolyForm Noncommercial License. Commercial use and managed deployments require a separate license.

The backlog is already waiting

Put your agent on it.

Build the habit your team will wish it had started six months ago.

Open Issopen