litellm/litellm-rust/crates/python-bridge/AGENTS.md
Yujong Lee d187fe8b2c refactor(rust): compose callback lifecycles in a chain and tighten both contracts
Compose the legacy and v1 lifecycles behind one PythonLifecycle through
LifecycleChain, built once in python-bridge, and simplify the boundary:

- drop LifecycleEvent::Started; begin carries the start time and is the first
  thing a lifecycle hears of a call
- give PythonLifecycle pass-through defaults for the threading steps and the
  stream notices, and document which steps thread and which observe
- export one is_cancellation from litellm-host-python
- callbacks-v1: private RequestFacts fields, a golden file per EventKind by
  construction, Python EVENTS and SUPPORTED_SCHEMAS pinned to the goldens
- callbacks-v1-python: the snapshot only parses (register is the only place
  that rejects), Handler owns call/is_async, Pending keeps a subscriber index
- callbacks-legacy-python: LegacyCallbacks becomes inherent PythonLogger
  methods, deployment hooks become free functions, retained headers are
  traversed and cleared
- define the shared terms in host-python/AGENTS.md and each contract's own
  terms in its crate
- remove test_loader.py, whose fork hook was never committed

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
2026-09-19 15:12:53 -07:00

8 KiB

  • Target invariants, not completion claims; these supersede the crate guidance below where they conflict
  • Keep this crate the product-specific PyO3 consumer of litellm-host-python
    • Own registration, input projection, the route host and the caller callables it answers operations with (file readers, token providers), public response/error construction and the per-call composition of machine, route host and callback contract
    • Legacy callback sharing (the caller's args, kwargs and request object, body/header roots, re-aliasing unchanged body keys) lives in litellm-callbacks-legacy-python behind PublicCall and LegacyPythonLifecycle; the bridge hands the public call over and keeps no copy
    • Routes never select a callback contract: callbacks::run_python_call is the only way a route starts a call, and src/callbacks.rs is the only place that names a contract crate
      • Every call runs the legacy Logging lifecycle, then the v1 lifecycle when anything is subscribed; with no subscriber the call is exactly the legacy-only call
      • A route describes itself once as a CallSurface; fields only the legacy contract reads are marked so and go when it goes
      • Retiring legacy for a route, or for all of them, is a change to that one composition, never to a route, the driver or an adapter
    • Value-oriented execution, sync waiting, nested-runtime checks, signal polling and panic containment live in litellm-host-python; native async work uses pyo3-async-runtimes, Serde output uses Pythonized<T>
    • Core owns typed native state, the route machine, provider preparation/I/O and normalization; the host driver owns terminal events; the legacy adapter in litellm-callbacks-legacy-python owns Logging dispatch policy
    • Python, Rust SDK and gateway use one lifecycle-bearing core route entrypoint; provider helpers stay private, never bridge-accessible transport drivers
    • Built-in provider/config/secret/auth/document preparation stays in Rust; caller-authored callbacks and focused Python-file reads run only at core-selected points
  • Target GIL-enabled CPython explicitly with #[pymodule(gil_used = true)]; detach Rust-only work
    • Free-threading requires separate runtime/concurrency validation; omitting the attribute does not opt out on PyO3 0.28+
  • Preserve public argument binding and Python object provenance
    • Project only consumed fields at reference read points; no eager whole-graph serialization or equality-based alias reconstruction
    • Preserve provider-specific upload/submission/poll observation and encoding boundaries; signed/build-captured bytes must not be silently reserialized
  • Conversion errors and every failure after the call starts are terminal
    • Disabled/unavailable native execution may select legacy once; callback exceptions never authorize fallback or replay
  • Use one ordinary inline async def driver in litellm/rust_bridge/lifecycle.py, with the native handle and call driver in litellm-host-python
    • Contract: start, resume_value, resume_error, idempotent close; explicitly tagged Await/Complete preserve awaitable final values
    • Validate Created/Running/Suspended/Closed protocol states; the machine yields ops, the driver emits one terminal event, the adapter chooses dispatch policy
    • Defer effectful setup/context reads/timestamps until start; unstarted-handle destruction releases inputs independently of Python finally
    • Catch only the selected await's errors; start/resume errors propagate, GeneratorExit closes without further awaits
    • Inline hooks preserve caller task/thread/loop and context writes; into_future creates a separate task and cannot satisfy this contract
  • Finalize fallible public response/error construction, replacements and metadata before terminal dispatch
  • Make ownership safe across suspension, re-entry, cancellation and GC
    • Keep native provider state typed in core; do not shuttle it through opaque Python transport/response classes
    • Prefer one retained Py<PyBaseException> via PyErr::into_value(py); reconstruct transient PyErrs, preserving identity, traceback, cause and context
    • Traverse every owned Python edge, including duplicate references; traversal cannot call Python
    • Take state out and mark Running under a short borrow, release borrows/locks before Python invocation, publish terminal state before finalizer-capable drops
    • Close/GC/deferred release are idempotent and re-entry-safe, including during Rust unwinding; release only owned references, never clear caller containers or mask the selected error
    • The machine owns its in-flight provider future; interrupt drops it synchronously, so provider captures are released before the driver returns and no task outlives the call
  • Verify behavior through a fresh, provenance-checked installed extension and positive native execution evidence before replacing the custom coroutine
    • Cover admitted provider workflows, binding/read-point/identity behavior, failure continuation, finalization, no replay, deferred gates, re-entry, GC and cancellation termination
    • Measure real conversion/copy costs before optimizing; preserve input contracts and capture lifetimes with PyBackedBytes, and lookup timing when interning names
    • Ship accurate _native.pyi declarations and typing markers; distinguish Future-returning bindings from coroutine-returning bindings
  • References: ownership, GC, exception transfer, re-entry

Rules for litellm-rust/crates/python-bridge.

Responsibility

python-bridge is the PyO3 boundary between Python LiteLLM and Rust transforms. Keep this crate thin. It exposes LiteLLM Rust APIs, assembles domain requests, maps domain errors to Python exceptions, and delegates generic conversion and GIL handling to litellm-host-python.

Bridge Shape

  • Prefer one stable method per top-level LiteLLM route, for example messages(...), calling the matching litellm-core entrypoint.
  • Do not add one exported PyO3 function per provider helper unless there is a measured reason.
  • Provider dispatch belongs in the litellm-core route module (e.g. litellm_core::messages), not in this PyO3 crate.
  • Python owns rollout state and fallback. Rust should return errors; Python decides whether to raise or fall back. For a rust-only provider/route (no Python reference), the Python side is a thin dispatch that calls Rust and raises when the bridge is unavailable, with no fallback.
  • Keep the Python interface minimal (well under 100 lines per route): it only marshals inputs and calls Rust. Do not add per-route feature flags, and do not put provider dispatch in litellm/main.py; it lives in a thin dispatch class under litellm/llms/<provider>/<route>/.

Data Handling

  • OCR payloads can contain personal data and large base64 images. Do not log payloads or provider responses.
  • Avoid copying large payloads more than needed. The current JSON round-trip is acceptable for the first scaffold, but future performance work should evaluate direct PyO3 conversion before expanding Rust coverage to image-heavy paths.
  • Do not expose raw Rust errors that include document contents or upstream bodies.

Tests

  • cargo test --workspace must compile this crate.
  • Python tests must cover bridge disabled, bridge enabled, and module-missing fallback behavior for every exposed route.