mirror of
https://github.com/pewdiepie-archdaemon/odysseus.git
synced 2026-10-08 07:52:20 +02:00
feat(runtime): resolve compact-runtime context window at turn preparation
The compact (clean v3) runtime had no effective context window: it learned a limit only reactively from a provider 400/413 and its terminal metrics carried no context_length. PR #41 addressed the reporting gap by probing provider metadata between the last model byte and [DONE], unauthenticated, and folded known-table and endpoint evidence into one "known" flag. Resolve the window once, before the first model request, instead: - src/agent_runtime/context_resolution.py adds a typed ContextResolution (effective value, evidence class, source, all observations, conflicts, provider_io, cached, secret-free probe errors). Evidence classes stay distinct: runtime_confirmed (llama.cpp /slots, /props, or a limit the provider stated this turn), provider_advertised (models catalog), operator_declared (client_runtime_context.model_context_window), known_table, unknown (0, never a default). - Selection is deterministic: runtime beats provider beats table; an operator declaration caps measured evidence and replaces weaker evidence. Disagreements are recorded as conflicts; a declaration below a measured value is a cap, above it a contradiction. - The provider probe forwards the turn's credentials only to the provider's own origin, runs URL resolution off the event loop, is bounded by one deadline, never raises, and caches remote results per credential fingerprint (shorter TTL for failures; local servers are re-probed). - stream_preview resolves at preparation (or accepts a supplied resolution), seeds the proactive trim budget from it when evidence is not unknown, and terminal metrics report only the stored resolution plus any limit the provider stated during the turn. Metrics perform no discovery. src/agent_loop.py and the regular runtime's legacy model_context probe are unchanged. A conftest guard keeps tests that drive the compact runtime with placeholder endpoints from performing real DNS/HTTP lookups.
This commit is contained in:
@@ -201,3 +201,36 @@ def _no_leaked_module_stubs():
|
||||
"at teardown.",
|
||||
pytrace=False,
|
||||
)
|
||||
|
||||
|
||||
@pytest.fixture(autouse=True)
|
||||
def _no_context_window_network_probe(request):
|
||||
"""Keep turn preparation from probing real provider metadata in tests.
|
||||
|
||||
The compact runtime resolves its context window before the first model
|
||||
request. Tests that drive it with placeholder endpoints must not perform
|
||||
DNS or HTTP lookups; modules that exercise the probe opt in with a
|
||||
module-level ``CONTEXT_PROBE_NETWORK = True`` and supply their own client.
|
||||
"""
|
||||
if getattr(request.module, "CONTEXT_PROBE_NETWORK", False):
|
||||
yield
|
||||
return
|
||||
try:
|
||||
from src.agent_runtime import context_resolution
|
||||
except Exception:
|
||||
yield
|
||||
return
|
||||
|
||||
async def _disabled_probe(endpoint_url, model, headers, is_local, observations, errors, timeout):
|
||||
errors.append("probe_disabled_in_tests")
|
||||
|
||||
# A private patcher keeps the shared ``monkeypatch`` fixture's teardown
|
||||
# order unchanged for tests that check their own sys.modules hygiene.
|
||||
patcher = pytest.MonkeyPatch()
|
||||
patcher.setattr(context_resolution, "_probe", _disabled_probe)
|
||||
context_resolution.clear_probe_cache()
|
||||
try:
|
||||
yield
|
||||
finally:
|
||||
patcher.undo()
|
||||
context_resolution.clear_probe_cache()
|
||||
|
||||
Reference in New Issue
Block a user