feat(server): add OpenAI Chat stream projections and hooks

[Pouyanpi]

Summary

Implements the stream surface already declared by the public OpenAI Chat guard contract, with staged projection/classifier modules and handwritten native error framing.

What Changed

• Adds reviewed guarded-delta, metadata, provider-error, and terminal-event classifications.
• Keeps unknown content-bearing deltas closed and choice indices strictly integer zero.
• Tests field policy, event inventory, provenance, transport declarations, and the hook binding against the maintained guard contract.
• Reuses the public request binding and contract without duplicating their earlier changes.

Review Notes

Review the declared event inventory, unsupported content channels, strict JSON, and native SSE error framing. Endpoint/HTTP integration lands in the final PR.

AI Assistance

• No AI tools were used.
• AI tools were used (Codex assisted with commit reconstruction, HTTP/error reconciliation, regression tests, and PR drafting). Human review of the new diff is pending.

Checklist

• I've read the CONTRIBUTING guidelines.
• This is maintainer-directed work.
• The PR title follows the project commit convention.
• Applicable declarations and tests are included.
• Generated changelog files were not edited manually.
• Automated and human review comments are addressed or answered.
• The responsible reviewer or team is mentioned.

Stack Position

Part 2 of 4.

• Previous: feat(server): execute guarded provider streams #2434
• Next: feat(server): bind guarded streams to HTTP #2436

Stack Context

Adds OpenAI Chat streaming on top of #2414 in four independently tested slices: provider-neutral execution, OpenAI stream modules, injected HTTP lifecycle handling, and integration with the existing buffered request pipeline. Concrete outbound HTTP, deployment configuration, and replacement execution remain separate work.

Please review each PR against its parent branch, not directly against the root base branch, except for part 1.

Order	PR	Branch	Base
1	#2434	pouyanpi/transparent-proxy-streaming-kernel-1	pouyanpi/openai-chat-buffered-integration-4
2	#2435	pouyanpi/openai-chat-streaming-contract-2	pouyanpi/transparent-proxy-streaming-kernel-1
3	#2436	pouyanpi/transparent-proxy-streaming-http-3	pouyanpi/openai-chat-streaming-contract-2
4	#2437	pouyanpi/openai-chat-streaming-integration-4	pouyanpi/transparent-proxy-streaming-http-3

[greptile-apps]

Confidence Score: 5/5

[Medium risk] Adds OpenAI stream validation and projection logic.

The PR appears safe to merge; no new in-scope issue was identified, and both previous findings remain resolved.

Summary

This PR adds OpenAI Chat stream projections, event classification, terminal-state hooks, and tests aligned with the guard contract.

• Closes unreviewed choice and delta fields while preserving declared metadata.
• Classifies guarded text, provider errors, metadata, and the terminal sentinel.
• Frames native SSE errors and validates stream completion.

Diagram

%%{init: {'theme': 'neutral'}}%%
flowchart LR
  A[Provider SSE event] --> B[Classify against stream projection]
  B --> C{Event role}
  C -->|Guarded text| D[Content check]
  C -->|Metadata or error| E[Stream hooks]
  D --> E
  E --> F{Valid terminal state?}
  F -->|Yes| G[Release event]
  F -->|No| H[Reject stream]

Loading

_{Reviews (3) · Last reviewed commit: "fix(server): validate OpenAI stream comp..." · Reviewed by Greptile}