fix(responses): accept assistant output_text messages without id/status in input

[MatejKosec]

Summary

• Make OutputMessage.id and OutputMessage.status optional (Option<> with #[serde(default)]) so clients can replay assistant turns in conversation history without including output metadata fields
• Fixes 400 deserialization errors when clients send output_text content in assistant messages without id and status (these are output-only metadata, not relevant for input context)
• All construction sites (stream converter, response builder) continue to set both fields explicitly — no change to output behavior
• Backwards compatible: payloads that include id and status still deserialize correctly

Context

An external regression report identified that assistant messages with output_text content fail with 400 (InputParam deserialization error) on the nvidia_dynamo endpoint. The root cause is that OutputMessage requires id and status as mandatory fields, but clients replaying multi-turn history naturally omit these.

The serde untagged enum cascade (InputItem → Item::Message → MessageItem::Output) fails at OutputMessage due to missing required fields, then fails all remaining variants, producing a generic deserialization error from axum's Json extractor before handler code even runs.

Test plan

• Added 4 serde unit tests covering:
  • Assistant output_text message without id/status (issue feat: nixl metadata store and retrieved from etcd #6 repro)
  • Full multi-turn conversation with output_text history
  • Reasoning items in input array (issue CI: Cache image on main branch #7 repro)
  • Backwards compat: OutputMessage with id and status still works
• CI: cargo test -p dynamo-async-openai
• CI: cargo test -p dynamo-llm (protocol conversion tests)

Summary by CodeRabbit

• Bug Fixes
  • Enhanced system flexibility by making message ID and status fields optional. The system now handles a broader range of message formats, including input replay scenarios and batch operations. This improvement strengthens integration capabilities with diverse external systems while maintaining full backward compatibility.

[coderabbitai]

Walkthrough

OutputMessage fields id and status are converted from required to optional types. The id field changes from String to Option, and status changes from OutputStatus to Option. Serde attributes enable deserialization when these fields are absent. All construction sites are updated to wrap values in Some(...).

Changes

Cohort / File(s)	Summary
Core Type Definition lib/async-openai/src/types/responses/response.rs	Made OutputMessage.id and OutputMessage.status optional with serde(default, skip_serializing_if = "Option::is_none"). Added test modules to validate deserialization with and without these fields.
Protocol Implementation lib/llm/src/protocols/openai/responses/mod.rs, lib/llm/src/protocols/openai/responses/stream_converter.rs	Updated OutputMessage construction to wrap id and status values in Some(...) across multiple locations (initialization, completion, final assembly). Updated tests to align with Option-wrapped field patterns.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Poem

🐰 With Option's gentle grace we dance,
Fields now flexible, given their chance,
Some(...) wraps both id and state,
Deserialization's now first-rate!

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately and concisely describes the main change: making OutputMessage id/status optional to accept assistant messages without these fields in input.
Description check	✅ Passed	The description covers all required template sections with sufficient detail: Overview (summary of the fix), Details (what changed and why), Where to start (specific files affected), and Related Issues (references the regression report context).
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

Tip

Try Coding Plans. Let us write the prompt for your AI agent so you can ship faster (with fewer bugs).
Share your feedback on Discord.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Caution

Some comments are outside the diff and can’t be posted inline due to platform limitations.

⚠️ Outside diff range comments (1)

lib/async-openai/src/types/responses/response.rs (1)

1548-1566: ⚠️ Potential issue | 🟡 Minor

Stale doc comment on MessageItem should be updated to reflect optional fields.

The MessageItem doc comment on lines 184–185 still reads:

OutputMessage: role=assistant, required id & status fields

This is now incorrect — both id and status are optional. The comment drives incorrect expectations for future readers trying to understand the disambiguation strategy.

📝 Proposed fix

-/// - OutputMessage: role=assistant, required id & status fields
+/// - OutputMessage: role=assistant (optional id & status — required in model output, optional in input history)

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@lib/async-openai/src/types/responses/response.rs` around lines 1548 - 1566,
Update the stale doc comment that describes OutputMessage/MessageItem to reflect
that the id and status fields are optional: change any text stating "required id
& status fields" or similar to indicate that id: Option<String> and status:
Option<OutputStatus> are optional, and adjust phrasing around role
(AssistantRole) and content (Vec<OutputMessageContent>) if needed; locate the
comment near the OutputMessage struct / MessageItem description and ensure it
accurately documents the optionality of id and status so readers understand the
disambiguation strategy.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Outside diff comments:
In `@lib/async-openai/src/types/responses/response.rs`:
- Around line 1548-1566: Update the stale doc comment that describes
OutputMessage/MessageItem to reflect that the id and status fields are optional:
change any text stating "required id & status fields" or similar to indicate
that id: Option<String> and status: Option<OutputStatus> are optional, and
adjust phrasing around role (AssistantRole) and content
(Vec<OutputMessageContent>) if needed; locate the comment near the OutputMessage
struct / MessageItem description and ensure it accurately documents the
optionality of id and status so readers understand the disambiguation strategy.

---

ℹ️ Review info

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between c18b475 and 2e5634c.

📒 Files selected for processing (3)

• lib/async-openai/src/types/responses/response.rs
• lib/llm/src/protocols/openai/responses/mod.rs
• lib/llm/src/protocols/openai/responses/stream_converter.rs