the-pair
Open-source AI pair programming desktop app: a read-only Mentor and an Executor agent cross-check each other's code to catch AI mistakes before they land. Bring any model, Claude, GPT, Gemini, Kimi, GLM, Qwen or local Ollama. No agent CLI needed. macOS / Windows / Linux.
Runssource: GitHubhomepageRustApache-2.0commit 332bc2be80d5
Rust, Apache-2.0 licensed. The project labels itself: agent orchestration, agentic coding, ai agents, ai code review, ai coding, ai coding agent, ai pair programming and claude code.
the-pair runs. An Argusic agent installed it in 50 minutes and hit 5 errors and fixed every one without help on a clean machine with no GPU, and the whole session was recorded.
All npm test suites pass (537 JS + 187 UI + 849 Rust = 1573 tests). TypeScript typechecks pass. Vite build succeeds. 9/10 headless e2e-web tests pass (1 flaky keyboard focus test). The Tauri desktop binary does not launch in this headless container due to missing webkit2gtk runtime libraries not installable without root.
What the agent ended up with on a clean machine, in its own words. How this is measured
Measured by Argusic on a fresh machine every time. Every number links to its evidence. Argusic Score 100 of 100 (the mean of this project's run scores).
At a glance
Subject data from GitHub, linked at the top of this page, refreshed . Test data by Argusic (CC BY 4.0); every number links to a run page with the full log, the recording, and their sha256 hashes.
What was measured
From the run that decided this verdict. Every line is a recorded value, not an estimate.

What went wrong, in the agent's own words
The agent's own words, worst first, never rewritten into advice.
- Missing Tauri system libraries (libwebkit2gtk-4.1-dev, libgtk-3-dev, ~50 other -dev packages)30 minutes
- Rust toolchain not found10 minutes
- Node.js v18 installed but project requires >=22.223 minutes
- Static .a libraries in sysroot caused lld to prefer them over shared libs, producing undefined symbol errors3 minutes
- 1 of 10 headless e2e tests fails (command palette Cmd+K key shortcut not detected in headless Chrome)
- Install time
- 50 minutes
- Cold machine to finish
- 51 minutes
- Errors hit and fixed
- 5 hit, 5 fixed with no human help
- How the result was proved
- npm test exited 0; JS tests 537 pass, UI tests 187 pass, Rust tests 849 pass; npm run build succeeded; typecheck:web and typecheck:node both clean
- Model tokens used
- 1,306,802
- Exact commit tested
- 332bc2be80d5
- Container image
- sha256:cdd920bce783
What was not measured
The other half, so you can see exactly how far this evidence goes. A gap is printed as a gap and never filled in.
- Peak GPU memory
- this ran on a CPU machine, so there was no GPU to sample
- Behaviour over a long run
- each attempt uses the smallest documented settings and a fixed time budget, so this measures whether it runs, not how it holds up under load
The end of the recording
The last lines of the terminal session that decided this verdict, exactly as they were recorded.
- Downloaded ~100+ Tauri system `.deb` packages (libwebkit2gtk-4.1-dev, libgtk-3-dev, etc.) via `ap… - Removed static `.a` library files that conflicted with shared library linking **Verification summary:** - `npm run test:js`: 537 pass, 0 fail - `npm run test:ui`: 187 pass, 0 fail - `npm run test:rust` (`cargo test`): 849 pass, 0 fail - `npm run typecheck:web`: clean exit - `npm run typecheck:node`: clean exit - `npm run build` (vite): builds successfully - `npm run e2e:web`: 9 passed, 1 flaky fail (unrelated Cmd+K palette keyboard detection in CI headl… - `npm test` (full suite): all three layers pass - The built vite output and all 1,573 test cases execute correctly
Replay the whole session, every command from a clean machine to this point.
How it was tested
One agent, one container, no human help. It installs the project the way its own documentation says to, runs it, and fixes what breaks. Everything below is recorded as it happened: the terminal session, the log and the exact commit. The full method.
Strengths and limits
Measured facts, not opinions. How this is written.
What went well
- Reached a running state on a clean machine, with the session recorded.
- Recovered from all 5 errors without a human stepping in, which says the failures are documented well enough to solve.
- Ran without a GPU, so it does not need one to start.
- Licensed Apache-2.0, as reported by its host.
What did not, or is not known
- Took 50 minutes to install, slower than the median of the 20 comparable projects Argusic has measured.
- Measured once so far. A single attempt shows it can run, not that it runs reliably.
Of the 21 agent-orchestration projects Argusic has installed and timed, the-pair was the 20th fastest to reach a running state, and 15 of 21 reached one at all.
What it is a reasonable choice for
- Trying it on a laptop or a small server: it reached a running state without a GPU.
Also tested, in the same area
Every one of these was installed and run by Argusic on a clean machine. Nothing appears here that was not tested.
Run history
Topics (from GitHub)
agent-orchestrationagentic-codingai-agentsai-code-reviewai-codingai-coding-agentai-pair-programmingclaude-codecode-reviewcodexcoding-agentdesktop-appdeveloper-toolsmulti-agentollamaopencodepair-programmingrusttaurivibe-coding
Embed the badge
Markdown for the project README. It links back here; terms on the terms page.
[](https://argusic.com/subject/the-pair)Questions
- Does the-pair run?
- Yes. the-pair runs. Argusic installed and launched it on a clean machine in 50 minutes, hitting 5 errors on the way, and recorded the session.
- How did Argusic test the-pair?
- On a fresh, disposable machine, with every command recorded and the repository pinned at commit 332bc2be80d5. One attempt is recorded, and the full method is on the methodology page.
- What did it cost to test the-pair?
- The run that produced this verdict cost $0.67: real compute and model cost, not a list price per million tokens. It is recorded on that run's page.
- How long does the-pair take to install?
- 50 minutes in the run that decided this verdict, from a clean machine to a running state. That is one measured attempt, not an average.
- Does the-pair need a GPU?
- Not to start. Argusic reached a running state on a machine with no GPU.
- What goes wrong when installing the-pair?
- 5 things broke in the recorded run, and 5 were fixed without human help. Each one, and the time it cost, is listed on this page.
- How does the-pair compare with the alternatives?
- Of the 21 agent-orchestration projects Argusic has installed and timed, the-pair was the 20th fastest to reach a running state, and 15 of 21 reached one at all.
- Where is the evidence for the-pair?
- The recorded run is on this page, each linking to its full log and terminal recording, stored with a sha256 fingerprint so it cannot be quietly altered.