QitOS is the torch-flavor framework for agent researchers.
Prototype methods, run benchmarks, and inspect long-horizon trajectories on one AgentModule + Engine kernel with built-in qita observability.
QitOS core is the small framework. Product-grade applications and showcase agents live in qitos-zoo, including planned apps such as qitos-coder and qitos-cyber-agent.
Quickstart · Tutorial Track · Benchmarks · CLI Reference · Changelog · Chinese README
QitOS is a research-first Agent framework: one AgentModule + Engine kernel,
with Session ownership, pause, durable recovery and fork. Extend tools,
providers, context, stores, sinks and sandboxes; inspect Trajectory through
read-only qita. Framework correctness does not guarantee arbitrary model task success.
-
Self-contained web tutorials: complete notes-Agent code, bilingual learning path and source-checked core API reference; tests execute the files shown on each page.
-
Master fixes Python 3.10 publication, repeated journal parsing and portable historical evidence verification; these are separately tested successors to the historical G5 runtime.
-
G5 framework qualification passed; S4 local integration complete. Runtime identity:
717b4cf1b23f2ed252cd03234ffd8605038d9567. -
Bilingual docs converge on installation → project → configuration → Session → inspection → recovery/extension.
-
The default development branch is
master, with CI/docs checks on pushes and PRs. Publication remains explicit. -
Documentation/tutorial qualification is recorded separately. Remote synchronization is verified. Docs CI passed; successor CI stabilization is tracked with exact results in the CI plan. Package releases and documentation deployments are tracked separately from CI qualification.
pip install qitos selects a published PyPI version, not an identity for unreleased G5.
Follow Installation, then the
credential-free Quickstart.
Real providers use agent.yaml, CredentialRef and an explicit resolver.
Docker file tools fail closed when unavailable; they do not fall back to host execution.
- Eight learning units: custom Agents, parallel tools, Sessions, context/memory, sandbox/artifacts, multi-agent work, qita and third-party extensions.
- Complete teaching files and example directory.
- Migration, limits and troubleshooting.
- Contributing and architecture.
- CHANGELOG, historical engineering progress and G5 evidence.
The historical 2663 passed / 50 skipped result belongs only to the runtime SHA
above, on Python 3.12.7. Advanced AgentModule.run() and historical trace
compatibility remain supported; they are not a second beginner path.