refactor(server): remove the unused V1 Codex session runtime

[juliusmarminge]

makeCodexSessionRuntime (the V1 Codex session runtime, ~2,700 lines) has had no production caller since the V2 Codex adapter took over. It was kept alive by one integration test and by CodexAdapterV2 importing two MCP elicitation helpers from it.

What changed

• Moved: describeMcpElicitation and toMcpElicitationResponse (plus their private schemas and helpers) now live in apps/server/src/provider/CodexMcpElicitation.ts, next to the other shared Codex modules (CodexDeveloperInstructions.ts, CodexToolPresentation.ts). The code is moved as-is. CodexAdapterV2 imports from the new module.
• Moved tests: the 12 MCP elicitation tests go to CodexMcpElicitation.test.ts. The 8 developer/browser instruction tests, which only exercised buildCodexDeveloperInstructions, go to CodexDeveloperInstructions.test.ts. Both are unchanged.
• Deleted:
  • CodexSessionRuntime.ts
  • The V1-only tests in CodexSessionRuntime.test.ts: thread history/pagination, buildTurnStartParams, hasConfiguredMcpServer, the memory-consolidation filter, codexSessionAppServerArgs, isRecoverableThreadResumeError, openCodexThread, plus one test that only restated an effect-codex-app-server error message.
  • CodexCollabRuntime.integration.test.ts, which booted the V1 runtime against a mock peer.
  • CodexCollabWire.test.ts, which tested the V1 child-notification router.
  • codexCollabMockPeer.cmd: only that integration test spawned it on Windows. The registry test that still uses the peer skips Windows.
  • codexSessionAppServerArgs, which only V1 used.
• Kept: codexCollabMockPeer.sh/.mjs and codexMultiAgentWire.json. ProviderInstanceRegistryLive.test.ts still spawns the peer as a fake codex for the readiness probe. collabAgentToolCall handling in CodexAdapterV2 is not touched: Codex 0.156.1 still emits v1 collab items.
• Small follow-ons: buildCodexInitializeParams is now private (knip flagged it as an unused export once V1 was gone). The manual-Effect-runner lint ceiling for the deleted test file is removed. The V2 TODO and the lineage doc now describe the missing paginated rollback path directly instead of pointing at V1.

Not ported (V2 behaves differently)

• Memory-consolidation thread suppression. V1 hid notifications from Codex's internal memory_consolidation subagent threads. V2 has no equivalent filter.
• Fallback to thread/start when resume fails recoverably. V1 started a fresh thread when thread/resume failed with a "thread not found / no rollout" style error. V2 resumeThread surfaces the error instead.
• Currency-symbol skill aliases. V1 buildTurnStartParams rewrote skill mentions typed with any currency sigil (€review, £review) into Codex's canonical $review form (fix(skills): support unicode currency symbols as skill aliases #12098). This PR deletes its only test, "sends currency skill aliases in Codex's canonical dollar form". V2 already lacked the rewrite before this PR; fix(server): Codex V2 runs skills typed with any currency sigil #13461 fixes that separately.
• Surfacing Codex stderr errors. V1 reported ERROR-level lines from Codex's stderr as session errors, skipping known-harmless ones. V2 reads stderr and discards it (packages/effect-codex-app-server/src/client.ts:272).

Verification

• vp exec tsc --noEmit -p . in apps/server: no error TS.
• vp lint on every touched TS file: 0 errors. The 2 existing no-unused-vars warnings in CodexAdapterV2.ts were already there before this change.
• vp fmt on touched files.
• knip --workspace apps/server --workspace packages/effect-codex-app-server --exports and knip --include files,dependencies: clean after making buildCodexInitializeParams private.
• vp test run on CodexMcpElicitation.test.ts, CodexDeveloperInstructions.test.ts, codexLaunchArgs.test.ts, CodexProvider.test.ts, CodexAdapterV2.test.ts, ProviderInstanceRegistryLive.test.ts: 6 files, 152 tests passed.
• git grep over apps, packages, scripts, docs, .github and root config for every deleted symbol and file name: no remaining references.
• Not run: the full server suite, repo-wide typecheck/lint (CI owns those), or any live Codex session.

Model: Claude Opus 5.5 (Claude Code)

🤖 Generated with Claude Code

---

[macroscopeapp]

Approvability

Verdict: Approved at 5c8928d

Macroscope's review found this PR approvable — This PR removes an unused V1 Codex runtime and its dedicated test harness while relocating shared MCP elicitation helpers without changing their implementation or the active V2 path. The remaining changes are limited to tests, tooling, documentation, and private-symbol cleanup.

^{You can add or adjust custom eligibility rules. Learn more.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ No successful main baseline artifact is available yet. This run establishes the initial measurement.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	—	4.9 KiB	—	6.8 KiB	✅
Codex	Thread snapshot wire	—	3.7 KiB	—	4.9 KiB	✅
Codex	Live turn WebSocket wire	—	1.1 KiB	—	2.0 KiB	✅
Codex	Live turn WebSocket decoded	—	20.4 KiB	—	29.3 KiB	✅
Codex	Live turn messages	—	1	—	8	✅
Claude	Total thread wire	—	4.9 KiB	—	6.8 KiB	✅
Claude	Thread snapshot wire	—	3.7 KiB	—	4.9 KiB	✅
Claude	Live turn WebSocket wire	—	1.2 KiB	—	2.0 KiB	✅
Claude	Live turn WebSocket decoded	—	20.8 KiB	—	29.3 KiB	✅
Claude	Live turn messages	—	2	—	8	✅

Baseline: unavailable · PR result: 5c8928d · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 106.1 KiB
• Claude decoded thread snapshot: 106.4 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.