proof surface
Taskmaster in the cockpit
Structured work tracking stays inside the same environment as coding, navigation, and runtime operation.
Terminal-first / local-first / operator-native
A terminal-first AI workspace for serious agentic coding.
AOC keeps context, memory, tasks, and runtime control inside one disciplined workspace so work can continue across sessions, machines, and real projects without collapsing into disposable chat tabs.
01
Terminal-first AI workspace
02
Local-model friendly and low-overhead
03
Structured, strategic agentic coding
04
Built from lived daily development practice
Why AOC matters
continuity engineAI-assisted development breaks down when context is stateless, decisions disappear, and work tracking lives in separate surfaces. AOC exists to keep the machine aligned with the project itself.
Operator signals
context in repo
memory persisted
tasks active
Live cockpit preview
Actual AOC session with explorer, runtime, pulse, and work surface
Explorer
repo-awareYazi file navigation / repo tree / active selection
Agent
runningPI runtime / project context / active task thread
Pulse
livesession telemetry / model path / operator state
Taskmaster
syncedwork queue / subtasks / current execution lane
Why AOC
A different thesis from browser-chat coding
AOC is not another glossy AI coding GUI. It is for people who want a machine-collaboration workflow that stays portable, disciplined, and grounded in the terminal.
No messy GUIs
AOC stays close to the terminal and the project itself, so the interface supports work instead of interrupting it.
No online requirement
AOC is compatible with local-first setups and can work with local models, giving developers more control over where and how they operate.
No subscription logic
The value is in the workflow, continuity, and operator discipline—not in locking serious coding behind cloud product mechanics.
Workflow contrast
From stateless prompting to structured operator flow.
browser-chat mode
- Context disappears between sessions
- Tasks live somewhere else
- Architecture decisions are easy to lose
- The machine behaves like a disposable assistant
AOC mode
- Context is persisted in project artifacts
- Tasks stay inside the same cockpit
- Memory survives across active work cycles
- The machine operates as a strategic coding partner
Product proof
Built from real daily development practice
AOC is not speculative portfolio fiction. The public site should visibly reflect the workspace, repo, and operating discipline behind Intrface development.
signal 01
Persistent project context, memory, and task state stored in-reposignal 02
Zellij-based cockpit layout with Yazi, PI runtime, Pulse, and Taskmastersignal 03
Built from real daily use as the development environment behind Intrface workProof gallery
Actual operator surfaces
proof surface
Repo-aware navigation
Yazi exposes the real project structure, including .aoc and task artifacts, instead of hiding the machine behind a browser layer.
Choose your path
Use AOC, learn the method, or bring it into team practice
The AOC public site should serve distinct intents without flattening them into the same CTA.
path
Developers
For people who want agentic coding to feel strategic, structured, and grounded in the machine they actually control.
Continue →path
Learners
For people who want to learn modern human-machine coding interaction through serious tooling rather than prompt theater.
Continue →path
Teams
For organizations exploring continuity, memory, operator control, and local-first adoption paths for coding agents.
Continue →Next step
Explore the product, learn the method, or work with Intrface directly.
AOC is both a real environment and a serious approach to human-machine coding. Choose the path that matches your intent and keep the workflow grounded in the machine.