OpenClicky / AI Bridge
Experimental orchestration between coding agents and computer-use controls.
A testbed for handing work between coding agents while sharing conversational state and controlling a real browser and screen.
01 Overview
What this is
OpenClicky and AI Bridge explore the seam between agents that write code and agents that operate a computer. The interesting part is not a single agent but the handoff: what context moves, who is allowed to click, and how a failure is recovered.
An experimental environment. Some workflows work; reliability is still the open question.
02 What it does
Capabilities and features
- Codex worker handoffs
- Shared conversational state across agents
- Browser and screen controls
- Native dragging and clicking
- Cross-agent development workflows
- Reliability improvements between runs
Described from the project's own notes. Unless a claim appears under Evidence, it has not been independently measured or benchmarked here.
03 Build notes
Recorded during development
- Presented as an ongoing experiment, not a finished orchestration product.
- Reliability work is described qualitatively in project notes; no pass-rate or benchmark is published.
04 Evidence
Only what can be pointed at
No link, artefact or measurement for OpenClicky / AI Bridge is published here yet. That is an honest empty state, not an oversight — the capability list above comes from project notes.
05 Limits
What this does not claim
- Reliability varies by workflow; no automated evaluation or pass rate is published.
- No public demo, installer or source link is available.
No screenshots or demos have been supplied for this project, so none are shown. The artwork above is generated concept art, labelled as such.
06 Status & next
Where it stands
Project notes; no dated public changelog yet.
Not published.
This site is independent. It is not connected to SOURCE infrastructure, Zenova production systems or any private account data.