agentHost/claude: Phase 6 — sendMessage, single-turn, no tools#314216
Open
TylerLeonhardt wants to merge 7 commits intomainfrom
Open
agentHost/claude: Phase 6 — sendMessage, single-turn, no tools#314216TylerLeonhardt wants to merge 7 commits intomainfrom
TylerLeonhardt wants to merge 7 commits intomainfrom
Conversation
- roadmap.md: mark Phase 4 as DONE, link merged PR #313780. - phase4-plan.md: record live-system smoke completion in §7.8; disabled-gate run skipped (covered by unit tests + env-var guard). - claudeAgent.test.ts: drop gratuitous 'as unknown as' cast in the CCAModel fixture (literal already matches CCAModelBilling exactly; plan §7.4 forbids unsafe casts in tests).
Handoff plan for Phase 5 (replace 7 throwing stubs in claudeAgent.ts). Locked against post-PR-#313841 reality (provisional sessions, onDidMaterializeSession, 30s empty-session GC) and the IAgent contract on origin/main. Decisions captured: - Non-fork createSession is synchronous and in-memory; fork deferred to Phase 6 (throws TODO). - IClaudeAgentSdkService surface mirrors IAgent (no dir parameter on listSessions); SDK loader caches resolved module, retries on failure, logs once. - listSessions joins SDK enumeration with workbench session DB metadata via ISessionDataService; per-entry try/catch resilience. - shutdown() routes per-session teardown through the same SequencerByKey<string> used by disposeSession() so concurrent shutdown/disposeSession cannot double-dispose a wrapper in Phase 6. - 14 unit tests defined (12 lifecycle + 2 resolved-config), including log-once contract and shutdown/disposeSession race guard.
Lands the ClaudeAgent IAgent provider behind the 'chat.agentHost.claudeAgent.enabled' setting (env gate VSCODE_AGENT_HOST_ENABLE_CLAUDE=1). Pins @anthropic-ai/claude-agent-sdk@0.2.112 in workspace + remote/. Implemented in this phase: * createSession - non-fork, in-memory wrapper only. Honors config.session for restore. The fork path and SDK session creation are deferred to Phase 6. * listSessions - SDK is source of truth; per-session DB read is a best-effort overlay (failure never excludes an entry). * disposeSession / shutdown - routed through a per-session SequencerByKey to serialize teardown. * getDescriptor, getProtectedResources, models, onDidSessionProgress, setClientCustomizations, setClientTools, onClientToolCallComplete, setCustomizationEnabled, authenticate, respondTo*Request - minimal Phase-5 wiring. Stubbed for Phase 6 (throw async 'TODO: Phase 6'): sendMessage, abortSession, changeModel, getSessionMessages, plus the createSession fork path. Tests: 29 unit tests in claudeAgent.test.ts cover the createSession restore-id path, listSessions overlay resilience, dispose serialization, and stub surfaces. Note: provisional / onDidMaterializeSession is intentionally omitted in Phase 5 (see plan section 3.3.1) - the workbench needs an immediate sessionAdded until the agent has real materialization work, which arrives in Phase 6 alongside SDK query() startup.
Implements the Phase 6 plan: provisional sessions materialize on first sendMessage, route a single-turn prompt through the Anthropic Claude Agent SDK's WarmQuery, and stream SDKMessages back as protocol AgentSignals via a pure mapSDKMessageToAgentSignals reducer. Tools remain denied (canUseTool: 'deny'); fork moves to Phase 6.5; Plan Mode UI moves to Phase 7. Highlights: - ClaudeAgent.sendMessage routes through _sessionSequencer to collapse concurrent first sends into one materialize + N ordered sends. - _materializeProvisional has two abort gates (post-startup + post-customizationDirectory write) so disposeSession landing mid-materialize cannot leak a WarmQuery subprocess. - ClaudeAgentSession owns the prompt iterator + per-turn deferreds; mapSDKMessageToAgentSignals is a pure reducer with state owned by the session. - IClaudeAgentSdkService gains startup() alongside listSessions(). Tests: 43 unit + 2 proxy-backed integration. Council-review fixes (C1 dispose race, C2 missing integration test, S1 cwd-less ratification) included.
# Conflicts: # package-lock.json # package.json # remote/package-lock.json # remote/package.json
Contributor
There was a problem hiding this comment.
Pull request overview
Implements Phase 6 of the agent-host Claude provider by introducing a provisional→materialized session lifecycle and a single-turn sendMessage path that materializes the Claude SDK on first prompt, then streams SDK messages through a pure reducer into protocol AgentSignals (with tools still denied for this phase).
Changes:
- Adds provisional session materialization on first
sendMessage, plus astartup()surface onIClaudeAgentSdkServiceto pre-warm the SDK subprocess. - Introduces
ClaudeAgentSession(query owner + message processing loop), and pure helpers for prompt/attachment resolution and SDK-message→signal mapping. - Adds a proxy-backed integration test and updates smoke/roadmap/plan docs for Phase 6 validation.
Show a summary per file
| File | Description |
|---|---|
| src/vs/platform/agentHost/test/node/claudeAgent.integrationTest.ts | New proxy-backed integration tests exercising the end-to-end pipeline with a stubbed CAPI and recording SDK service. |
| src/vs/platform/agentHost/node/claude/smoke.md | Updates smoke procedure and validation checks for Phase 6+ streaming/materialization behavior. |
| src/vs/platform/agentHost/node/claude/roadmap.md | Marks Phase 4 as done and links to the landed PR. |
| src/vs/platform/agentHost/node/claude/phase6-plan.md | Adds detailed Phase 6 implementation plan and acceptance criteria (reference documentation). |
| src/vs/platform/agentHost/node/claude/phase5-plan.md | Adds Phase 5 plan documentation (reference documentation). |
| src/vs/platform/agentHost/node/claude/phase4-plan.md | Updates Phase 4 smoke checklist status/results. |
| src/vs/platform/agentHost/node/claude/claudePromptResolver.ts | New pure helper to convert prompt + attachments into Anthropic content blocks. |
| src/vs/platform/agentHost/node/claude/claudeMapSessionEvents.ts | New pure mapper from SDK messages to agent-host protocol actions/signals. |
| src/vs/platform/agentHost/node/claude/claudeAgentSession.ts | New session runtime that owns the SDK query, prompt iterable, abort, and message loop. |
| src/vs/platform/agentHost/node/claude/claudeAgentSdkService.ts | New lazy SDK wrapper service with listSessions() + Phase 6 startup(). |
| src/vs/platform/agentHost/node/claude/claudeAgent.ts | Implements provisional sessions, materialization, real sendMessage, env wiring, and session config schema. |
| src/vs/platform/agentHost/node/agentHostServerMain.ts | Registers IClaudeAgentSdkService for server-mode agent host. |
| src/vs/platform/agentHost/node/agentHostMain.ts | Registers IClaudeAgentSdkService for local/utility-process agent host. |
| src/vs/platform/agentHost/common/claudeSessionConfigKeys.ts | New well-known Claude session config keys/types (permissionMode). |
| remote/package.json | Adds @anthropic-ai/claude-agent-sdk dependency to remote bundle. |
| package.json | Adds @anthropic-ai/claude-agent-sdk dependency. |
| package-lock.json | Lockfile updates for the new dependency and its transitive deps. |
| eslint.config.js | Allows importing @anthropic-ai/claude-agent-sdk in the agentHost area. |
Copilot's findings
Files not reviewed (1)
- remote/package-lock.json: Language not supported
- Files reviewed: 18/20 changed files
- Comments generated: 2
… abort) Two Copilot-reviewer comments on #314216: 1. listSessions: wrap _sdkService.listSessions() in try/catch. AgentService.listSessions fans out across providers via Promise.all; an SDK dynamic-import failure would otherwise nuke every other provider's session list. Now logs and returns []. 2. dispose: abort _provisionalSessions AbortControllers before super.dispose(). Previously a racing first sendMessage parked inside _writeCustomizationDirectory could pass the materialize abort gates and call _sessions.set on a disposed DisposableMap, orphaning the WarmQuery. Aborting first triggers the existing post-customization-write abort gate, which asyncDisposes the WarmQuery. Tests: 2 new regressions (listSessions empty on SDK throw; agent.dispose() during racing materialize disposes the WarmQuery). 45/45 unit + 2/2 integration pass.
aeschli
previously approved these changes
May 4, 2026
dmitrivMS
approved these changes
May 4, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Implements the Phase 6 plan for the Claude
IAgentprovider: provisional sessions materialize on firstsendMessage, route a single-turn prompt through the Anthropic Claude Agent SDK'sWarmQuery, and streamSDKMessages back as protocolAgentSignals via a puremapSDKMessageToAgentSignalsreducer.Tools remain denied (
canUseTool: 'deny') for this phase. Fork moves to Phase 6.5 (needs message-UUID lookup); plan-mode UI to Phase 7.What's in this stack
This PR stacks 4 commits on
main:agentHost/claude: post-Phase-4 cleanup— pre-merged separatelyagentHost/claude: lock Phase 5 implementation plan— Phase 5 planagentHost/claude: Phase 5 — IAgent provider skeleton— Phase 5 implementationagentHost/claude: Phase 6 — sendMessage, single-turn, no tools— this workHighlights of the Phase 6 commit
ClaudeAgent.sendMessageroutes through_sessionSequencerso concurrent first sends collapse into one materialize + N ordered sends — no double-fork of the SDK subprocess._materializeProvisional. Post-sdk.startup()and post-_writeCustomizationDirectory()checks ensure that adisposeSessionlanding mid-materialize disposes the producedWarmQueryinstead of leaking a subprocess into_sessions. (Council-review C1.)ClaudeAgentSessionowns the prompt iterator + per-turn deferreds;mapSDKMessageToAgentSignalsis a pure reducer with state owned by the session — independently testable, no shared mutable state between sessions.IClaudeAgentSdkService.startup()added alongsidelistSessions()so the agent can be tested with a fake SDK without spawning a real subprocess.SessionResponsePartis allocated oncontent_block_startBEFORE any delta can arrive — deltas are SDK-ordered after the start — so the protocol's "part-must-precede-delta" invariant (actions.ts:233, 540) holds by construction.Tests
claudeAgent.test.tscovering: materialize lifecycle (provisional/abort/double-materialize), prompt iterator pumping, dispose ↔ shutdown ↔ send race matrix, the C1 dispose-during-customization-write regression, mapper output for text + thinking + result, and the SDK loader's cache-and-log-once contract.claudeAgent.integrationTest.ts(council-review C2): realClaudeProxyService(loopback HTTP server), stubbed Copilot CAPI streaming cannedMessageStreamEvents — exercises the full agent → proxy → CAPI → SSE → agent pipeline including theANTHROPIC_AUTH_TOKENnonce contract.Validation
tsgo --noEmit -p src/tsconfig.json: cleaneslinton all modified files: cleannpm run gulp -- hygiene: clean for affected filesOut of scope (deferred per plan)
canUseTool: 'deny'(Phase 6.5+)config.fork(Phase 6.5)getSessionMessages(Phase 13)