docs(plan-execute): clarify Responses history requirement

[ryan-lempka]

Document that plan/execute needs conversation history, including tool calls and results. Explain that Responses requests using only previous_response_id or a provider conversation ID can miss the edit and stay on the planner.

Apps embedding libsy must supply that history before routing. This documents the limitation and the full-history workaround instead of adding automatic history replay in #897.

This limitation affects Responses clients that send continuation IDs instead of conversation history. With Switchyard’s supplied Codex configuration, Codex includes the conversation history in each request, so this limitation does not apply.

[github-actions]

PR Preview Action v1.8.1
🚀 View preview at https://NVIDIA-NeMo.github.io/Switchyard/pr-preview/pr-905/
Built to branch gh-pages at 2026-10-02 17:43 UTC. Preview will be ready when the GitHub Pages deployment is complete.

[coderabbitai]

Navigate logical layers of code changes, visualize relationships, and explore their blast radius.

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository: NVIDIA-NeMo/Switchyard/.coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 128a33c5-2568-4630-986d-e7fb88e774bc

📥 Commits

Reviewing files that changed from the base of the PR and between 16cbe59 and 4090e65.

📒 Files selected for processing (2)

• docs/getting_started.md
• docs/routing_algorithms/plan_execute_routing.md

Included review availability: This review used your included allowance. Your plan provides up to 12 included reviews per hour; 10 remain after this review.

---

Walkthrough

The documentation now describes the conversation history required for plan/execute routing, including tool calls and results. It explains that continuation IDs alone do not provide this history and updates the “Optional settings” label to a Markdown heading.

Changes

Plan/execute conversation history

Layer / File(s)	Summary
Document history requirements docs/getting_started.md, docs/routing_algorithms/plan_execute_routing.md	The guides state that callers must provide the earlier conversation history, including tool calls and results. The routing guide explains that continuation IDs alone do not provide the history, and that session IDs cannot detect edits missing from the request. It also formats “Optional settings” as a Markdown heading.

Priority: ⬇️ Low

Estimated code review effort: 2 (Simple) | ~7 minutes

Merge Risk: ⚪ Minimal · up to 4090e

The documented full-history guidance matches how plan/execute detects tool activity. No issue identified here prevents merging after normal checks.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 0…
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely describes the main change: documenting the Responses API history requirement for plan/execute routing.

• Autopilot · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Autopilot is currently an internal CodeRabbit preview.

---

_{A rabbit reads the history trail,
With tool calls tucked beneath its tail.
The router finds each step in view,
The heading stands up straight and true.
Then hops away through docs anew.}

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