fix(auto-reply): preserve sessions after compaction failures (#70479)

Summary:
- The PR removes the auto-reply compaction-failure session reset hook, adds preserved-session recovery guidance for overflow/compaction failure paths, and updates focused tests, docs, and the changelog.
- Reproducibility: yes. at source level with high confidence. Current main routes both embedded overflow paylo ... resetSessionAfterCompactionFailure, and the PR body includes before/after terminal proof of those branches.

Automerge notes:
- PR branch already contained follow-up commit before automerge: fix(auto-reply): drop dead compaction reset hook
- PR branch already contained follow-up commit before automerge: fix(auto-reply): preserve sessions after compaction failures

Validation:
- ClawSweeper review passed for head 193d3c0fdd.
- Required merge gates passed before the squash merge.

Prepared head SHA: 193d3c0fdd
Review: https://github.com/openclaw/openclaw/pull/70479#issuecomment-4325128777

Co-authored-by: FullerStackDev <263060202+fuller-stack-dev@users.noreply.github.com>
Co-authored-by: vincentkoc <25068+vincentkoc@users.noreply.github.com>
Co-authored-by: clawsweeper[bot] <274271284+clawsweeper[bot]@users.noreply.github.com>
Approved-by: takhoffman
Co-authored-by: takhoffman <781889+takhoffman@users.noreply.github.com>

CHANGELOG.md

@@ -4929,7 +4929,8 @@ Docs: https://docs.openclaw.ai
 - Providers/SDK retry: cap long `Retry-After` sleeps in Stainless-based Anthropic/OpenAI model SDKs so 60s+ retry windows surface immediately for OpenClaw failover instead of blocking the run. (#68474) Thanks @jetd1.
 - Agents/TTS: preserve spoken text in TTS tool results while defusing reply directives in transcript content, so future turns remember voice replies without treating spoken `MEDIA:` or voice tags as delivery metadata. (#68869) Thanks @zqchris.
 - Providers/OpenAI: harden Voice Call realtime transcription against OpenAI Realtime session-update drift, forward language and prompt hints, and add live coverage for realtime STT.
-- Agents/Pi embedded runs: suppress the "⚠️ Agent couldn't generate a response" warning when the assistant already delivered user-visible content through a messaging tool and the turn ended cleanly (`stopReason=stop`). Real failure modes (tool errors, provider `stopReason=error`, interrupted tool use) still surface the existing "verify before retrying" warning. Fixes #70396. (#70425) Thanks @neeravmakwana.
+- Agents/Pi embedded runs: suppress the "⚠️ Agent couldn't generate a response" warning when the assistant already delivered user-visible content through a messaging tool and the turn ended cleanly (`stopReason=stop`). Real failure modes (tool errors, provider `stopReason=error`, interrupted tool use) still surface the existing "verify before retrying" warning. Fixes #70396. (#70425) Thanks @neeravmakwana.
+- Auto-reply/WebChat: preserve the active session mapping when context-overflow recovery or auto-compaction fails, and return retry, `/compact`, and `/new` guidance instead of silently rotating to a fresh session. Fixes #70472. (#70479) Thanks @fuller-stack-dev.
 - Gateway/Linux: wrap gateway-managed supervisor, PTY, MCP stdio, and browser child processes in a tiny `/bin/sh` shim that raises the child's own `oom_score_adj` on Linux, so under cgroup memory pressure the kernel prefers transient workers over the long-lived gateway. Opt out with `OPENCLAW_CHILD_OOM_SCORE_ADJ=0`. Fixes #70404. (#70419) Thanks @neeravmakwana.
 - Providers/Moonshot: stop strict-sanitizing Kimi's native tool_call IDs (shaped like `functions.<name>:<index>`) on the OpenAI-compatible transport, so multi-turn agentic flows through Kimi K2.6 no longer break after 2-3 tool-calling rounds when the serving layer fails to match mangled IDs against the original tool definitions. Adds a `sanitizeToolCallIds` opt-out to the shared `openai-compatible` replay family helper and wires Moonshot to it. Fixes #62319. (#70030) Thanks @LeoDu0314.
 - Dependencies/security: override transitive `uuid` to `14.0.0`, clearing the runtime advisory across dependencies.

docs/reference/session-management-compaction.md

@@ -274,6 +274,11 @@ In the embedded Pi agent, auto-compaction triggers in two cases:
 number of tokens`, `input token count exceeds the maximum number of input
 tokens`, `input is too long for the model`, `ollama error: context length
 exceeded`, and similar provider-shaped variants) → compact → retry.
+   If overflow recovery still fails, OpenClaw surfaces explicit guidance to the
+   user and preserves the current session mapping instead of silently rotating
+   the session key to a fresh session id. The next step is operator-controlled:
+   retry the message, run `/compact`, or run `/new` when a fresh session is
+   preferred.
 2. **Threshold maintenance**: after a successful turn, when:
 
 `contextTokens > contextWindow - reserveTokens`

src/auto-reply/reply/agent-runner-execution.test.ts

@@ -25,6 +25,10 @@ const state = vi.hoisted(() => ({
   isCliProviderMock: vi.fn((_: unknown) => false),
   isInternalMessageChannelMock: vi.fn((_: unknown) => false),
   createBlockReplyDeliveryHandlerMock: vi.fn(),
+  isCompactionFailureErrorMock: vi.fn((_: string | undefined) => false),
+  isContextOverflowErrorMock: vi.fn((_: string | undefined) => false),
+  isLikelyContextOverflowErrorMock: vi.fn((_: string | undefined) => false),
+  updateSessionStoreMock: vi.fn(),
 }));
 
 const GENERIC_RUN_FAILURE_TEXT =
@@ -87,10 +91,11 @@ vi.mock("../../agents/pi-embedded-helpers.js", () => ({
     }
     return undefined;
   },
-  isCompactionFailureError: () => false,
-  isContextOverflowError: () => false,
+  isCompactionFailureError: (message?: string) => state.isCompactionFailureErrorMock(message),
+  isContextOverflowError: (message?: string) => state.isContextOverflowErrorMock(message),
   isBillingErrorMessage: () => false,
-  isLikelyContextOverflowError: () => false,
+  isLikelyContextOverflowError: (message?: string) =>
+    state.isLikelyContextOverflowErrorMock(message),
   isOverloadedErrorMessage: (message: string) => /overloaded|capacity/i.test(message),
   isRateLimitErrorMessage: (message: string) =>
     /rate.limit|too many requests|429|usage limit/i.test(message),
@@ -101,7 +106,7 @@ vi.mock("../../agents/pi-embedded-helpers.js", () => ({
 vi.mock("../../config/sessions.js", () => ({
   resolveGroupSessionKey: vi.fn(() => null),
   resolveSessionTranscriptPath: vi.fn(),
-  updateSessionStore: vi.fn(),
+  updateSessionStore: state.updateSessionStoreMock,
 }));
 
 vi.mock("../../globals.js", () => ({
@@ -270,10 +275,13 @@ function createFollowupRun(): FollowupRun {
 function createMockReplyOperation(): {
   replyOperation: ReplyOperation;
   failMock: ReturnType<typeof vi.fn>;
+  updateSessionIdMock: ReturnType<typeof vi.fn>;
 } {
   const failMock = vi.fn();
+  const updateSessionIdMock = vi.fn();
   return {
     failMock,
+    updateSessionIdMock,
     replyOperation: {
       key: "main",
       sessionId: "session",
@@ -282,7 +290,7 @@ function createMockReplyOperation(): {
       phase: "running",
       result: null,
       setPhase: vi.fn(),
-      updateSessionId: vi.fn(),
+      updateSessionId: updateSessionIdMock,
       attachBackend: vi.fn(),
       detachBackend: vi.fn(),
       complete: vi.fn(),
@@ -389,7 +397,6 @@ function createMinimalRunAgentTurnParams(overrides?: {
     shouldEmitToolResult: () => true,
     shouldEmitToolOutput: () => false,
     pendingToolTasks: new Set<Promise<void>>(),
-    resetSessionAfterCompactionFailure: async () => false,
     resetSessionAfterRoleOrderingConflict: async () => false,
     isHeartbeat: false,
     sessionKey: "main",
@@ -410,6 +417,19 @@ describe("buildContextOverflowRecoveryText", () => {
     expect(text).not.toContain("heartbeat model bleed");
   });
 
+  it("keeps the preserved-session copy with the existing overflow hint", () => {
+    const text = buildContextOverflowRecoveryText({
+      preserveSessionMapping: true,
+      cfg: {},
+      primaryProvider: "openrouter",
+      primaryModel: "qwen3.6-plus",
+    });
+
+    expect(text).toContain("kept this conversation mapped to the current session");
+    expect(text).toContain("reserveTokensFloor");
+    expect(text).not.toContain("reset our conversation");
+  });
+
   it("points to heartbeat model bleed when the last runtime model matches configured heartbeat.model", () => {
     const text = buildContextOverflowRecoveryText({
       cfg: {
@@ -593,6 +613,13 @@ describe("runAgentTurnWithFallback", () => {
     state.isInternalMessageChannelMock.mockReturnValue(false);
     state.createBlockReplyDeliveryHandlerMock.mockReset();
     state.createBlockReplyDeliveryHandlerMock.mockReturnValue(undefined);
+    state.isCompactionFailureErrorMock.mockReset();
+    state.isCompactionFailureErrorMock.mockReturnValue(false);
+    state.isContextOverflowErrorMock.mockReset();
+    state.isContextOverflowErrorMock.mockReturnValue(false);
+    state.isLikelyContextOverflowErrorMock.mockReset();
+    state.isLikelyContextOverflowErrorMock.mockReturnValue(false);
+    state.updateSessionStoreMock.mockReset();
     state.runWithModelFallbackMock.mockImplementation(async (params: FallbackRunnerParams) => ({
       result: await params.run("anthropic", "claude"),
       provider: "anthropic",
@@ -1250,7 +1277,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -1367,7 +1393,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -1443,7 +1468,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -1508,7 +1532,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -1569,7 +1592,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -1628,7 +1650,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -1681,7 +1702,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -1738,7 +1758,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -1785,7 +1804,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -1952,7 +1970,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks,
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -2004,7 +2021,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -2049,7 +2065,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -2338,7 +2353,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks,
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -2385,7 +2399,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks,
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -2431,7 +2444,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks,
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -2484,7 +2496,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks,
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -2715,7 +2726,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -2761,7 +2771,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -2826,7 +2835,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -2883,7 +2891,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -2944,7 +2951,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3032,7 +3038,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks,
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3167,7 +3172,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3251,7 +3255,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3298,7 +3301,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3344,7 +3346,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3393,7 +3394,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3450,7 +3450,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3515,7 +3514,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3579,7 +3577,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3637,7 +3634,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3690,7 +3686,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3742,7 +3737,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3782,7 +3776,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3843,7 +3836,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3892,7 +3884,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3946,7 +3937,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -3993,7 +3983,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4032,7 +4021,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4137,7 +4125,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4367,7 +4354,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4383,6 +4369,95 @@ describe("runAgentTurnWithFallback", () => {
     }
   });
 
+  it("preserves the active session when embedded overflow recovery fails", async () => {
+    state.isContextOverflowErrorMock.mockReturnValue(true);
+    state.runEmbeddedPiAgentMock.mockResolvedValueOnce({
+      payloads: [],
+      meta: {
+        error: {
+          message: "400 The prompt is too long: 203557, model maximum context length: 196607",
+        },
+      },
+    });
+
+    const activeSessionEntry = { sessionId: "session", updatedAt: 1 } as SessionEntry;
+    const activeSessionStore = { "agent:main:main": activeSessionEntry };
+    const { replyOperation, failMock, updateSessionIdMock } = createMockReplyOperation();
+    const runAgentTurnWithFallback = await getRunAgentTurnWithFallback();
+    const result = await runAgentTurnWithFallback({
+      ...createMinimalRunAgentTurnParams({
+        sessionCtx: {
+          Provider: "webchat",
+          MessageSid: "msg",
+        } as unknown as TemplateContext,
+      }),
+      replyOperation,
+      sessionKey: "agent:main:main",
+      getActiveSessionEntry: () => activeSessionEntry,
+      activeSessionStore,
+      storePath: "/tmp/sessions.json",
+    });
+
+    expect(result.kind).toBe("final");
+    if (result.kind === "final") {
+      expect(result.payload.text).toContain("kept this conversation mapped to the current session");
+      expect(result.payload.text).toContain("reserveTokensFloor");
+      expectRecordFields(requireRecord(getReplyPayloadMetadata(result.payload), "reply metadata"), {
+        deliverDespiteSourceReplySuppression: true,
+      });
+    }
+    expect(failMock).toHaveBeenCalledWith(
+      "run_failed",
+      expect.objectContaining({
+        message: "400 The prompt is too long: 203557, model maximum context length: 196607",
+      }),
+    );
+    expect(activeSessionStore["agent:main:main"]?.sessionId).toBe("session");
+    expect(updateSessionIdMock).not.toHaveBeenCalled();
+    expect(state.updateSessionStoreMock).not.toHaveBeenCalled();
+  });
+
+  it("preserves the active session when compaction failure is thrown before reply", async () => {
+    state.isCompactionFailureErrorMock.mockReturnValue(true);
+    state.runEmbeddedPiAgentMock.mockRejectedValueOnce(
+      new Error("Auto-compaction failed: nothing to compact"),
+    );
+
+    const activeSessionEntry = { sessionId: "session", updatedAt: 1 } as SessionEntry;
+    const activeSessionStore = { "agent:main:main": activeSessionEntry };
+    const { replyOperation, failMock, updateSessionIdMock } = createMockReplyOperation();
+    const runAgentTurnWithFallback = await getRunAgentTurnWithFallback();
+    const result = await runAgentTurnWithFallback({
+      ...createMinimalRunAgentTurnParams({
+        sessionCtx: {
+          Provider: "webchat",
+          MessageSid: "msg",
+        } as unknown as TemplateContext,
+      }),
+      replyOperation,
+      sessionKey: "agent:main:main",
+      getActiveSessionEntry: () => activeSessionEntry,
+      activeSessionStore,
+      storePath: "/tmp/sessions.json",
+    });
+
+    expect(result.kind).toBe("final");
+    if (result.kind === "final") {
+      expect(result.payload.text).toContain("kept this conversation mapped to the current session");
+      expect(result.payload.text).toContain("reserveTokensFloor");
+      expectRecordFields(requireRecord(getReplyPayloadMetadata(result.payload), "reply metadata"), {
+        deliverDespiteSourceReplySuppression: true,
+      });
+    }
+    expect(failMock).toHaveBeenCalledWith(
+      "run_failed",
+      expect.objectContaining({ message: "Auto-compaction failed: nothing to compact" }),
+    );
+    expect(activeSessionStore["agent:main:main"]?.sessionId).toBe("session");
+    expect(updateSessionIdMock).not.toHaveBeenCalled();
+    expect(state.updateSessionStoreMock).not.toHaveBeenCalled();
+  });
+
   it("surfaces gateway reauth guidance for known OAuth refresh failures", async () => {
     state.runEmbeddedPiAgentMock.mockRejectedValueOnce(
       new Error(
@@ -4407,7 +4482,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4447,7 +4521,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4485,7 +4558,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4525,7 +4597,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4565,7 +4636,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4601,7 +4671,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4636,7 +4705,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4675,7 +4743,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4742,7 +4809,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4797,7 +4863,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4882,7 +4947,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -4940,7 +5004,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -5003,7 +5066,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -5072,7 +5134,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -5143,7 +5204,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -5207,7 +5267,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",
@@ -5274,7 +5333,6 @@ describe("runAgentTurnWithFallback", () => {
       shouldEmitToolResult: () => true,
       shouldEmitToolOutput: () => false,
       pendingToolTasks: new Set(),
-      resetSessionAfterCompactionFailure: async () => false,
       resetSessionAfterRoleOrderingConflict: async () => false,
       isHeartbeat: false,
       sessionKey: "main",

src/auto-reply/reply/agent-runner-execution.ts

@@ -875,15 +875,18 @@ function resolveHeartbeatBleedHint(params: {
 
 export function buildContextOverflowRecoveryText(params: {
   duringCompaction?: boolean;
+  preserveSessionMapping?: boolean;
   cfg: FollowupRun["run"]["config"];
   agentId?: string;
   primaryProvider?: string;
   primaryModel?: string;
   activeSessionEntry?: SessionEntry;
 }): string {
-  const prefix = params.duringCompaction
-    ? "⚠️ Context limit exceeded during compaction. I've reset our conversation to start fresh - please try again."
-    : "⚠️ Context limit exceeded. I've reset our conversation to start fresh - please try again.";
+  const prefix = params.preserveSessionMapping
+    ? "⚠️ Auto-compaction could not recover this turn. I kept this conversation mapped to the current session. Please try again, use /compact, or use /new to start a fresh session."
+    : params.duringCompaction
+      ? "⚠️ Context limit exceeded during compaction. I've reset our conversation to start fresh - please try again."
+      : "⚠️ Context limit exceeded. I've reset our conversation to start fresh - please try again.";
   return (
     prefix +
     (resolveHeartbeatBleedHint({
@@ -1134,7 +1137,6 @@ export async function runAgentTurnWithFallback(params: {
   shouldEmitToolResult: () => boolean;
   shouldEmitToolOutput: () => boolean;
   pendingToolTasks: Set<Promise<void>>;
-  resetSessionAfterCompactionFailure: (reason: string) => Promise<boolean>;
   resetSessionAfterRoleOrderingConflict: (reason: string) => Promise<boolean>;
   isHeartbeat: boolean;
   sessionKey?: string;
@@ -1315,7 +1317,6 @@ export async function runAgentTurnWithFallback(params: {
   let fallbackProvider = params.followupRun.run.provider;
   let fallbackModel = params.followupRun.run.model;
   let fallbackAttempts: RuntimeFallbackAttempt[] = [];
-  let didResetAfterCompactionFailure = false;
   let didRetryTransientHttpError = false;
   let liveModelSwitchRetries = 0;
   let bootstrapPromptWarningSignaturesSeen = resolveBootstrapWarningSignaturesSeen(
@@ -2231,20 +2232,19 @@ export async function runAgentTurnWithFallback(params: {
       });
 
       // Some embedded runs surface context overflow as an error payload instead of throwing.
-      // Treat those as a session-level failure and auto-recover by starting a fresh session.
+      // Preserve the active session mapping and surface explicit guidance instead
+      // of silently rotating the session key to a new session id.
       const embeddedError = runResult.meta?.error;
-      if (
-        embeddedError &&
-        isContextOverflowError(embeddedError.message) &&
-        !didResetAfterCompactionFailure &&
-        (await params.resetSessionAfterCompactionFailure(embeddedError.message))
-      ) {
-        didResetAfterCompactionFailure = true;
+      if (embeddedError && isContextOverflowError(embeddedError.message)) {
+        defaultRuntime.error(
+          `Auto-compaction failed (${embeddedError.message}). Preserving existing session mapping for ${params.sessionKey ?? params.followupRun.run.sessionId}.`,
+        );
         params.replyOperation?.fail("run_failed", embeddedError);
         return {
           kind: "final",
           payload: markAgentRunFailureReplyPayload({
             text: buildContextOverflowRecoveryText({
+              preserveSessionMapping: true,
               cfg: runtimeConfig,
               agentId: params.followupRun.run.agentId,
               primaryProvider: params.followupRun.run.provider,
@@ -2364,18 +2364,17 @@ export async function runAgentTurnWithFallback(params: {
         };
       }
 
-      if (
-        isCompactionFailure &&
-        !didResetAfterCompactionFailure &&
-        (await params.resetSessionAfterCompactionFailure(message))
-      ) {
-        didResetAfterCompactionFailure = true;
+      if (isCompactionFailure) {
+        defaultRuntime.error(
+          `Auto-compaction failed (${message}). Preserving existing session mapping for ${params.sessionKey ?? params.followupRun.run.sessionId}.`,
+        );
         params.replyOperation?.fail("run_failed", err);
         return {
           kind: "final",
           payload: markAgentRunFailureReplyPayload({
             text: buildContextOverflowRecoveryText({
               duringCompaction: true,
+              preserveSessionMapping: true,
               cfg: runtimeConfig,
               agentId: params.followupRun.run.agentId,
               primaryProvider: params.followupRun.run.provider,

src/auto-reply/reply/agent-runner.ts

@@ -1437,12 +1437,6 @@ export async function runReplyAgent(params: {
           activeIsNewSession = true;
         },
       });
-    const resetSessionAfterCompactionFailure = async (reason: string): Promise<boolean> =>
-      resetSession({
-        failureLabel: "compaction failure",
-        buildLogMessage: (nextSessionId) =>
-          `Auto-compaction failed (${reason}). Restarting session ${sessionKey} -> ${nextSessionId} and retrying.`,
-      });
     const resetSessionAfterRoleOrderingConflict = async (reason: string): Promise<boolean> =>
       resetSession({
         failureLabel: "role ordering conflict",
@@ -1471,7 +1465,6 @@ export async function runReplyAgent(params: {
         shouldEmitToolResult,
         shouldEmitToolOutput,
         pendingToolTasks,
-        resetSessionAfterCompactionFailure,
         resetSessionAfterRoleOrderingConflict,
         isHeartbeat,
         sessionKey,