Commit Graph
10 Commits
Author SHA1 Message Date
jif 4885eb6c52 Retry transient Guardian sampling failures (#38997)
## Why

Guardian sampling can encounter temporary connection or service failures while
opening a WebSocket, starting a request, or consuming its response stream.

## What changed

- Retry transient transport, streaming, rate-limit, and server failures up to
  two times across the full sampling flow.
- Recover from unauthorized responses through the configured auth manager and
  discard pooled connections that may contain stale credentials.
- Allow sampler initialization to continue when its initial WebSocket prewarm
  fails, so the first sample can establish a fresh connection.

## Testing

Added coverage for recovery after prewarm failures and `503` responses, and for
stopping once the transient retry budget is exhausted.

GitOrigin-RevId: af74c2df95c4241184ca641ef4418db6aa4ab695
2026-08-17 12:45:50 +00:00
jif 37cf6c84c0 Include images in Guardian v2 transcripts (#38987)
## What changed

- Add the `features.guardianv2.transcript.include_images` opt-in setting.
- Include recent images from user and assistant messages and configured tool outputs in Guardian v2 sampling requests.
- Bound image context to the four most recent images and 8 MiB total, and omit image detail hints when constructing the sampling request.

## Testing

- Cover configuration parsing and image forwarding from messages and tool outputs.

GitOrigin-RevId: b0ae6b6182535fd084ea103ed01dc3c803d7d1f7
2026-08-17 10:56:13 +00:00
jif 742edd6c16 Prioritize new Guardian classifications under load (#38596)
## What changed

- Expand the Guardian sampling pool from 8 to 16 WebSocket connections.
- When the pool is full, supersede the oldest request that has already produced a score before replacing an unfinished classification. Treat superseded classifications as a no-op in the extension.
- Stop response WebSocket work when its event consumer is dropped, including while waiting for the connection lock or draining a completed sample.
- Allow retries across both initially warmed connections for retryable stream failures.

## Testing

- Add a concurrent sampler test that fills the pool and verifies scored drains are replaced before an unfinished classification.

GitOrigin-RevId: 9e4df7b01516094726ff3e11878f8ca742d21b03
2026-08-14 15:55:01 +00:00
felixxia-oai 08c48e3700 Reuse compatible parent compactions in Guardian V2 (#38564)
## Why

Guardian V2 classification requests should retain compacted conversation context only when the parent and classifier model configurations can share the encrypted compaction.

## What changed

- Resolve the classifier model's compaction compatibility hash when initializing its sampler.
- Add the latest encrypted parent compaction to classification input when both models have the same non-empty compatibility hash.
- Ignore missing, empty, or unidentified compactions, and omit parent context when compatibility cannot be established.

## Testing

- Cover latest-compaction selection and invalid compaction data.
- Cover matching, mismatched, missing, empty, and changed compatibility hashes.

GitOrigin-RevId: f0fcb09fd028cd2f397fdbc3d541b4bc464b0beb
2026-08-14 11:55:10 +00:00
felixxia-oai db5b4a4cfb Preserve Guardian transcript boundaries in sampling input (#38555)
## What changed

- Represent Guardian sampling input as an ordered list of text entries.
- Send each transcript entry and approval-request delimiter as a separate input
  content item while preserving the existing transcript byte limit.

## Testing

- Update transcript and sampler tests to verify entry boundaries, multi-item
  request serialization, and bounded transcript behavior.

GitOrigin-RevId: 433e2e6ca18ce9288025715f5956295bfe852e98
2026-08-14 11:02:01 +00:00
jif a44f266a4a Initialize Guardian V2 samplers per thread (#38553)
## What changed

- Register Guardian V2 for thread startup as well as tool-call lifecycle events.
- When `guardianv2` is enabled, create and store a thread-local Luna sampler using the thread's model provider, authentication policy, session metadata, originator, and service tier.
- Emit a warning if sampler initialization fails, and skip tool classification when the thread has no sampler.

## Testing

- Update the extension test to enable `guardianv2`, run thread startup, and verify tool-call sampling through the initialized connection.

GitOrigin-RevId: 5ab732e85c9827495c67f57117eb5be450a7f90e
2026-08-14 10:58:56 +00:00
Dylan Hurd 4eff3b788b Configure Guardian sampling for responses lite (#38527)
## What changed

- Opt Guardian V2 Luna sampler WebSocket connections into responses lite.
- Send `reasoning.context` as `all_turns` on sampling requests.

## Testing

- Extend sampler and extension tests to verify the connection header and reasoning context.

GitOrigin-RevId: 4b0263c01b8259ed5268c222dcf7559912baba8c
2026-08-14 08:16:23 +00:00
jif 66919805ea Pool Guardian sampling WebSocket connections (#38406)
## What changed

- Prewarm two Responses WebSocket connections and grow the pool on demand, up to eight connections, so overlapping samples can run concurrently.
- Lease each connection to a single sample and return healthy connections to the idle pool after completion.
- Replace closed or 55-minute-old connections and retry a sample once after a retryable or stream error.
- Keep the sampler available when the second prewarm connection cannot be opened.

## Testing

- Cover connection reuse, partial prewarm success, concurrent pool growth, and retrying an expired connection on another warm connection.

GitOrigin-RevId: 3ee7b05adc00a2135ba7cb35f6e4455199bf3784
2026-08-13 17:00:59 +00:00
jif 911012490c Return Luna samples when streamed JSON completes (#38383)
## Why

Structured output can be complete before the Responses stream emits its terminal
events. Waiting for those events unnecessarily delays the sampler and can leave it
waiting after usable output has arrived.

## What changed

- Return a Luna sample as soon as accumulated text deltas parse as a complete JSON
  object.
- Continue draining the remaining response events in the background so the
  authenticated WebSocket stays reusable for subsequent samples.
- Enforce the output-size limit while accumulating deltas on the early-return path.

## Testing

- Cover returning complete JSON without terminal response events.
- Verify that early return still allows the authenticated WebSocket to be reused.

GitOrigin-RevId: 80c9963f1540f7862356aaad3503e3071a8fb92a
2026-08-13 14:20:38 +00:00
jif a7b8c074b5 Add the Guardian V2 Luna sampler (#38368)
## What changed

- Add a `LunaSampler` that opens an authenticated Responses WebSocket and reuses it for structured, tool-free `gpt-5.6-luna` requests.
- Carry the host's provider, authentication, proxy, attribution, and service-tier configuration into the connection and requests.
- Require strict JSON schemas, preserve per-request reasoning effort and turn metadata, and reject missing or oversized output.

## Testing

- Add a WebSocket test covering authentication and attribution headers, connection reuse, structured request fields, per-turn metadata, reasoning effort, and returned output.

GitOrigin-RevId: aef85fd8ddcfcb7c76a9ede0d3191e3a0321656b
2026-08-13 12:35:56 +00:00