Fix context-engine compaction ownership for Codex sessions by joshavant · Pull Request #91590 · openclaw/openclaw

joshavant · 2026-06-09T03:51:10Z

Summary

keep owning context-engine compaction primary for queued budget compaction when a Codex/native harness is also selected
request Codex native thread/compact/start only as a bounded secondary follow-up after successful owning context-engine compaction
clear stale Codex thread-bootstrap projection metadata before that native follow-up mutates the bound thread
add focused regression coverage for the queued handoff, secondary Codex 4xx behavior, Discord /new stale model cleanup, and the real Codex app-server bridge path

Refs #90496. This addresses the live-reproduced context-engine/native-compaction breakage from the issue and adds direct regression coverage for two adjacent issue symptoms. It does not claim to close every reported symptom around real production Codex provider 4xxs, scheduled-session metadata recreation, or automatic transcript rotation policy.

Real behavior proof

Behavior addressed: Budget auto-compaction can be short-circuited by the Codex app-server native path before an owning context engine compacts the session.

Real environment tested: Final-head local runtime proof on b5db02fd5a1cd673a17b86cc92eef10792a2b82a, using the queued compaction entry point plus a proof harness that calls the real Codex app-server compact bridge with a fake client at the socket boundary. Live Discord lane was also tested with a Convex-leased discord credential and gateway started from this branch.

Exact steps or command run after this patch: pnpm tsx /private/tmp/openclaw-90496-final-head-proof.ts. Live Discord proof was run with 1Password-injected Convex broker credentials: op-with-service-account run --env-file /private/tmp/openclaw-90496-op.env -- pnpm tsx /private/tmp/openclaw-90496-discord-proof.ts.

Evidence after fix: Final-head proof output showed queued context-engine compaction succeeded and the secondary Codex bridge sent thread/compact/start with request:"after_context_engine":

{
  "sha": "b5db02fd5a1cd673a17b86cc92eef10792a2b82a",
  "compactResult": {
    "ok": true,
    "compacted": true,
    "reason": "proof-context-engine-compacted",
    "result": {
      "details": {
        "engine": "proof-context-engine",
        "nativeHarnessCompaction": {
          "ok": true,
          "compacted": false,
          "result": {
            "details": {
              "backend": "codex-app-server",
              "signal": "thread/compact/start",
              "pending": true,
              "request": "after_context_engine",
              "trigger": "budget"
            }
          }
        }
      }
    }
  },
  "bridgeRequests": [
    {
      "method": "thread/compact/start",
      "params": {
        "threadId": "thread-proof-final-head"
      }
    }
  ],
  "bindingAfter": {
    "contextEngine": {
      "schemaVersion": 1,
      "engineId": "proof-context-engine",
      "policyFingerprint": "proof-policy"
    }
  },
  "assertions": {
    "contextEngineCompacted": true,
    "contextEngineMarkerWritten": true,
    "secondaryCodexResultOk": true,
    "secondaryCodexRequestMarked": true,
    "secondaryCodexBridgeSignal": true,
    "outboundBridgeRequest": true,
    "projectionCleared": true,
    "didNotTakeOldNonManualSkip": true
  }
}

Observed result after fix: The owning context engine compacts first. Only after that succeeds, the secondary Codex bridge sends thread/compact/start with request:"after_context_engine", clears the stale context-engine projection marker, and does not take the old reason:"non_manual_trigger" skip path.

What was not tested live: A real production Codex provider provider_error_4xx, scheduled/reminder stale model metadata recreation, and a new automatic transcript-rotation product policy for genuinely unrecoverable compaction.

Added Issue-Shape Regression Proof

src/agents/embedded-agent-runner/compact.hooks.test.ts now covers a secondary Codex/native provider_error_4xx after successful context-engine compaction. The primary result remains ok:true, compacted:true, and the 4xx is nested under codexNativeCompaction instead of overriding the context-engine result.
src/auto-reply/reply/session.test.ts now covers the exact Discord channel shape from Discord channel remains trapped in oversized session after /new; compaction fails provider_error_4xx and model drifts from codex/gpt-5.5 to gpt-5.4 #90496: a stale codex/gpt-5.4 auto fallback from openai/gpt-5.5, oversized token counters, and /new. The reset creates a new session and clears provider/model fallback provenance plus stale token counters so the configured codex/gpt-5.5 default can apply.

Live Discord proof

The live Discord proof on this branch leased a discord credential from Convex, sent a real marker mention, observed the SUT bot reply, reset a seeded stale session entry, and invoked queued budget compaction for the Discord channel session.

Redacted live output:

{
  "credential": {
    "kind": "discord",
    "source": "convex",
    "id": "<redacted>"
  },
  "discord": {
    "guildId": "<redacted>",
    "channelId": "<redacted>",
    "sentMessageId": "<redacted>",
    "replyMessageId": "<redacted>",
    "markerObserved": true
  },
  "reset": {
    "rpcOk": true,
    "before": {
      "providerOverride": "codex",
      "modelOverride": "gpt-5.4",
      "fallbackOriginProvider": "openai",
      "fallbackOriginModel": "gpt-5.5"
    },
    "after": {}
  },
  "compaction": {
    "result": {
      "ok": true,
      "compacted": true,
      "reason": "proof-context-engine-compacted"
    },
    "markerWritten": true,
    "markerEvents": [
      {
        "sessionKey": "agent:qa:discord:channel:<redacted>",
        "target": "budget",
        "force": false
      }
    ]
  }
}

Regression proof

On current origin/main at 80f1ae6ffe, the same live Discord proof reproduced the narrowed failure: Discord marker observed, reset clean, but budget compaction returned compacted:false, reason:"codex app-server owns automatic compaction", and markerWritten:false.

Verification

.agents/skills/autoreview/scripts/autoreview --mode local passed with no accepted/actionable findings after the added 90496 edge-case tests.
node scripts/run-vitest.mjs src/agents/embedded-agent-runner/compact.hooks.test.ts src/auto-reply/reply/session.test.ts passed, 69/69 agent tests and 104/104 auto-reply session tests.
node scripts/run-vitest.mjs extensions/codex/src/app-server/compact.test.ts passed, 19/19 Codex bridge tests.
pnpm tsx /private/tmp/openclaw-90496-final-head-proof.ts passed on b5db02fd5a1cd673a17b86cc92eef10792a2b82a.
Live Discord proof passed on this branch with a Convex-leased discord credential.
git diff --check passed.

clawsweeper · 2026-06-09T03:53:15Z

Codex review: needs maintainer review before merge. Reviewed June 9, 2026, 10:14 PM ET / 02:14 UTC.

Summary
The PR makes context-engine-owned compaction run before a bounded internal Codex native compaction follow-up, guards Codex binding/projection updates, waits for active native turns, and adds focused regression coverage.

PR surface: Source +520, Tests +819. Total +1339 across 14 files.

Reproducibility: yes. by source inspection: current main runs Codex native harness compaction before the owning context engine and the Codex native path skips budget triggers, matching the PR body's live current-main failure proof.

Review metrics: 1 noteworthy metric.

Plugin SDK surface: 0 public SDK params added; 1 private Codex harness capability. The post-context-engine compaction hint stays internal, lowering plugin API compatibility risk before merge.

Merge readiness
Overall: 🐚 platinum hermit
Proof: 🦞 diamond lobster
Patch quality: 🐚 platinum hermit
Result: ready for maintainer review.

Overall follows the weaker of proof and patch quality, so missing proof can cap an otherwise strong patch.

Rank-up moves:

[P2] Get explicit Codex/session-state owner review for the binding/projection sequencing before merge.

Risk before merge

[P1] The PR changes when Codex binding projection metadata is cleared and when thread/compact/start runs; incorrect sequencing could desynchronize OpenClaw context-engine state from the native Codex thread.
[P1] The PR deliberately leaves remaining symptoms in Discord channel remains trapped in oversized session after /new; compaction fails provider_error_4xx and model drifts from codex/gpt-5.5 to gpt-5.4 #90496 outside scope, including production Codex provider 4xxs, scheduled metadata recreation, and automatic transcript-rotation policy.

Maintainer options:

Owner-review the compaction handoff (recommended)
Have a Codex/session-state owner verify the binding re-read, projection clear, and secondary native request ordering against live-session upgrade expectations before merge.

Next step before merge

[P2] Protected maintainer handling and session-state sequencing risk require explicit owner review rather than an automated repair PR.

Security
Cleared: No concrete security or supply-chain regression was found; the diff does not change workflows, dependencies, secrets storage, or package execution surfaces.

Review details

Best possible solution:

Land this only after a Codex/session-state owner accepts the binding/projection sequencing and keep the remaining symptoms tracked in #90496.

Do we have a high-confidence way to reproduce the issue?

Yes, by source inspection: current main runs Codex native harness compaction before the owning context engine and the Codex native path skips budget triggers, matching the PR body's live current-main failure proof.

Is this the best way to solve the issue?

Yes, likely: keeping the context engine primary and making Codex native compaction a private bounded follow-up is the narrowest owner-boundary fix I found, with maintainer review still needed for session-state sequencing.

AGENTS.md: found and applied where relevant.

Codex review notes: model gpt-5.5, reasoning high; reviewed against 9a1f2022b127.

Label changes

Label justifications:

P1: The linked behavior can leave a Discord agent channel trapped in an oversized session, breaking a real agent/channel workflow.
merge-risk: 🚨 session-state: The PR intentionally changes Codex thread binding, projection metadata, and context-engine/native compaction ordering.
rating: 🐚 platinum hermit: Overall readiness is 🐚 platinum hermit; proof is 🦞 diamond lobster and patch quality is 🐚 platinum hermit.
status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Sufficient (live_output): The PR body includes final-head runtime live output plus redacted live Discord proof showing the changed compaction behavior after the patch.
proof: sufficient: Contributor real behavior proof is sufficient. The PR body includes final-head runtime live output plus redacted live Discord proof showing the changed compaction behavior after the patch.

Evidence reviewed

PR surface:

Source +520, Tests +819. Total +1339 across 14 files.

View PR surface stats

Area	Files	Added	Removed	Net
Source	8	600	80	+520
Tests	6	823	4	+819
Docs	0	0	0	0
Config	0	0	0	0
Generated	0	0	0	0
Other	0	0	0	0
Total	14	1423	84	+1339

Acceptance criteria:

[P1] node scripts/run-vitest.mjs src/agents/embedded-agent-runner/compact.hooks.test.ts src/auto-reply/reply/session.test.ts.
[P1] node scripts/run-vitest.mjs extensions/codex/src/app-server/compact.test.ts extensions/codex/src/app-server/session-binding.test.ts extensions/codex/src/app-server/run-attempt.test.ts.
[P1] git diff --check.

What I checked:

Repository policy applied: Root and scoped OpenClaw review policy were read; the Codex/session-state change falls under the AGENTS.md guidance requiring whole-path review, Codex dependency inspection, and protected maintainer handling. (AGENTS.md:1, 9a1f2022b127)
Current main still has the old short-circuit: Current main calls native harness compaction before context-engine-owned compaction can run, so the PR is not obsolete on main. (src/agents/embedded-agent-runner/compact.queued.ts:264, 9a1f2022b127)
Current Codex native compaction skips non-manual triggers: Current main returns reason: "non_manual_trigger" for non-manual Codex app-server compaction, matching the reported budget-compaction failure shape. (extensions/codex/src/app-server/compact.ts:139, 9a1f2022b127)
PR makes context engine primary: The PR skips the native harness preflight when the context engine owns compaction, then runs the native harness only as an after_context_engine secondary follow-up after successful compaction. (src/agents/embedded-agent-runner/compact.queued.ts:290, a7d2bd2a5113)
Private Codex follow-up capability: The post-context-engine compaction request is kept behind a private internal harness capability rather than adding a public plugin SDK parameter. (src/agents/harness/compaction.ts:31, a7d2bd2a5113)
Binding/projection guard added: The PR re-reads the Codex binding under a per-binding lock before clearing context-engine projection metadata and sending thread/compact/start. (extensions/codex/src/app-server/compact.ts:212, a7d2bd2a5113)

Likely related people:

steipete: Git history shows repeated Codex app-server lifecycle/module work, including the app-server controls and lifecycle seam refactors that define this runtime boundary. (role: feature-history owner; confidence: high; commits: 31a0b7bd42a5, 8d72aafdbb8d, 3b65e2302a55; files: extensions/codex/src/app-server/run-attempt.ts, extensions/codex/src/app-server/compact.ts)
vincentkoc: Recent history on Codex app-server auth/startup and the latest stable release touched the same Codex runtime area involved in binding and auth-profile safety. (role: recent adjacent contributor; confidence: medium; commits: f1cc8f0cfc7c, 859eb0666282, 5181e4f7c82b; files: extensions/codex/src/app-server/run-attempt.ts, extensions/codex/src/app-server/session-binding.ts)
shakkernerd: Current-main blame in this shallow checkout points the central compaction and Codex binding lines at a recent commit by Shakker, making them a plausible routing candidate for current surface context. (role: recent area contributor; confidence: medium; commits: 56d201fa670b; files: src/agents/embedded-agent-runner/compact.queued.ts, extensions/codex/src/app-server/compact.ts, src/agents/harness/compaction.ts)
joshavant: Beyond authoring this PR, prior merged work by this contributor touches reply/session and auth-adjacent paths that overlap the linked Discord/session recovery symptoms. (role: adjacent contributor; confidence: low; commits: 731d4666d25f, 1769fb2aa1d6; files: src/auto-reply/reply/session.test.ts, src/agents/harness/compaction.ts)

What the crustacean ranks mean

🦀 challenger crab: rare, exceptional readiness with strong proof, clean implementation, and convincing validation.
🦞 diamond lobster: very strong readiness with only minor maintainer review expected.
🐚 platinum hermit: good normal PR, likely mergeable with ordinary maintainer review.
🦐 gold shrimp: useful signal, but proof or patch confidence is still limited.
🦪 silver shellfish: thin signal; proof, validation, or implementation needs work.
🧂 unranked krab: not merge-ready because proof is missing/unusable or there are serious correctness or safety concerns.
🌊 off-meta tidepool: rating does not apply to this item.

Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics.

How this review workflow works

ClawSweeper keeps one durable marker-backed review comment per issue or PR.
Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
Maintainers can also comment @clawsweeper review to request a fresh review only.
Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.

joshavant · 2026-06-09T05:28:46Z

@clawsweeper re-review

I updated the PR body with final-head proof on 2e24ceab886a1999d4dbc0d74854c8e3e27dcbbd:

queued context-engine compaction succeeds with ok:true, compacted:true, reason:"proof-context-engine-compacted"
the secondary Codex bridge path sends thread/compact/start
the secondary bridge result includes request:"after_context_engine" and trigger:"budget"
stale context-engine projection metadata is cleared before native Codex compaction
the old reason:"non_manual_trigger" skip path is not taken
live Discord primary proof was rerun on final head with a Convex-leased discord credential

clawsweeper · 2026-06-09T05:28:49Z

🦞🧹
ClawSweeper re-review requested.

I asked ClawSweeper to review this item again.
Action: item re-review queued (workflow sweep.yml, event repository_dispatch).
Result: the existing ClawSweeper review comment will be edited in place when the review finishes.

Re-review progress:

State: Complete
Detail: The targeted re-review finished, the durable review comment was updated, and the synced verdict was routed.
Run: https://github.com/openclaw/clawsweeper/actions/runs/27185919614
Updated: 2026-06-09T05:36:49.489Z

joshavant · 2026-06-09T07:27:46Z

Follow-up hardening pushed in fc4c870c57 for the session-state sequencing concern.

What changed:

The secondary after_context_engine Codex native compaction path now re-reads the Codex app-server binding under the binding mutation lock before clearing context-engine projection metadata or sending thread/compact/start.
Binding writes and clears now serialize through the same per-binding in-process queue plus the cross-process file lock. This prevents a same-process write/clear from slipping between compare, projection clear, and native compaction start.
The guarded native start is bounded to the binding-lock contract timeout. Peer binding mutations wait longer than that bounded RPC instead of timing out on the lock.
Pre-aborted guarded compaction skips before clearing projection, so abort cannot leave projection metadata cleared when Codex compaction never starts.
Missing binding clears keep the old no-op behavior and do not recreate deleted session directories.

Verification after this commit:

.agents/skills/autoreview/scripts/autoreview --mode local returned clean with no accepted/actionable findings.
node scripts/run-vitest.mjs extensions/codex/src/app-server/compact.test.ts extensions/codex/src/app-server/session-binding.test.ts src/agents/embedded-agent-runner/compact.hooks.test.ts passed after autoreview.
CI on fc4c870c57 is green, including Real behavior proof.

…91590) * fix(agents): keep context-engine compaction primary * fix(codex): request native compaction after context engines * test: cover 90496 compaction and reset edge cases * fix(codex): guard secondary native compaction binding * fix(codex): keep native compaction hint internal * fix(codex): wait for active native turns before resume

…26.6.6) (#1040) This PR contains the following updates: | Package | Update | Change | |---|---|---| | [ghcr.io/openclaw/openclaw](https://openclaw.ai) ([source](https://github.com/openclaw/openclaw)) | patch | `2026.6.5` → `2026.6.6` | --- ### Release Notes <details> <summary>openclaw/openclaw (ghcr.io/openclaw/openclaw)</summary> ### [`v2026.6.6`](https://github.com/openclaw/openclaw/blob/HEAD/CHANGELOG.md#202666) [Compare Source](openclaw/openclaw@v2026.6.5...v2026.6.6) ##### Highlights - Security boundaries are substantially tighter across transcripts, sandbox binds, host environment inheritance, MCP stdio, Codex HTTP access, native search policy, elevated sender checks, deleted-agent ACP bypasses, loopback tools, Discord moderation, and Teams group actions; exec approvals now fail closed on timeout. ([#91529](openclaw/openclaw#91529), [#91618](openclaw/openclaw#91618), [#91615](openclaw/openclaw#91615), [#91619](openclaw/openclaw#91619), [#91741](openclaw/openclaw#91741), [#91745](openclaw/openclaw#91745), [#91746](openclaw/openclaw#91746), [#91748](openclaw/openclaw#91748), [#91749](openclaw/openclaw#91749), [#91750](openclaw/openclaw#91750), [#91751](openclaw/openclaw#91751), [#91752](openclaw/openclaw#91752), [#91763](openclaw/openclaw#91763), [#89938](openclaw/openclaw#89938)) Thanks [@joshavant](https://github.com/joshavant), [@pgondhi987](https://github.com/pgondhi987), [@mmaps](https://github.com/mmaps), [@eleqtrizit](https://github.com/eleqtrizit), [@shakkernerd](https://github.com/shakkernerd), and [@drobison00](https://github.com/drobison00). - Telegram delivery is safer and more coherent: account-scoped topics route to the right agent, streamed text survives tool calls, `/compact` works on generic ingress, callback handling uses concrete APIs, draft chunking is shared, durable dispatch dedupe moved into the SDK, and unauthorized DM text stays out of cache and prompt context. ([#91189](openclaw/openclaw#91189), [#88682](openclaw/openclaw#88682), [#89588](openclaw/openclaw#89588), [#90212](openclaw/openclaw#90212), [#91876](openclaw/openclaw#91876), [#91874](openclaw/openclaw#91874), [#91904](openclaw/openclaw#91904), [#91478](openclaw/openclaw#91478), [#91915](openclaw/openclaw#91915)) Thanks [@codysai001](https://github.com/codysai001), [@alexzhu0](https://github.com/alexzhu0), [@joelnishanth](https://github.com/joelnishanth), [@snowzlm](https://github.com/snowzlm), [@obviyus](https://github.com/obviyus), and [@sallyom](https://github.com/sallyom). - iMessage recovery and delivery now cover always-on inbound restart, durable echo markers, block streaming, idle approval discovery, hardened outbound transport, and actionable inbound startup diagnostics. ([#91335](openclaw/openclaw#91335), [#91449](openclaw/openclaw#91449), [#88969](openclaw/openclaw#88969), [#88530](openclaw/openclaw#88530), [#91783](openclaw/openclaw#91783), [#91785](openclaw/openclaw#91785)) Thanks [@omarshahine](https://github.com/omarshahine), [@jmissig](https://github.com/jmissig), and [@colmbrogan](https://github.com/colmbrogan). - Browser and MCP connectivity gained existing-session CDP support, discovered WebSocket validation, default-profile `cdpUrl` handling, safer browser-output boundaries, Streamable HTTP loopback transport, corrected OAuth/SSE authorization handling, and broader schema compatibility. ([#91422](openclaw/openclaw#91422), [#89851](openclaw/openclaw#89851), [#91736](openclaw/openclaw#91736), [#91747](openclaw/openclaw#91747), [#91451](openclaw/openclaw#91451), [#80143](openclaw/openclaw#80143)) Thanks [@pgondhi987](https://github.com/pgondhi987), [@anagnorisis2peripeteia](https://github.com/anagnorisis2peripeteia), [@lifuyue](https://github.com/lifuyue), [@eleqtrizit](https://github.com/eleqtrizit), [@LiuwqGit](https://github.com/LiuwqGit), and [@HemantSudarshan](https://github.com/HemantSudarshan). - Control UI startup and first-reply latency are lower through cached model metadata, removal of the startup catalog wait, lazy slash-command loading, and first-event tracing with slow-reply diagnostics. ([#91531](openclaw/openclaw#91531), [#91538](openclaw/openclaw#91538), [#91568](openclaw/openclaw#91568), [#91583](openclaw/openclaw#91583), [#91598](openclaw/openclaw#91598)) - Provider support expands with OpenRouter OAuth onboarding and Claude Fable 5 adaptive thinking, while Codex sessions keep correct compaction ownership, local models skip guardian review, dynamic tool progress normalizes cleanly, and Gemma 4 reasoning replay is preserved. ([#91830](openclaw/openclaw#91830), [#91882](openclaw/openclaw#91882), [#91590](openclaw/openclaw#91590), [#88630](openclaw/openclaw#88630), [#88768](openclaw/openclaw#88768), [#91696](openclaw/openclaw#91696)) Thanks [@Patrick-Erichsen](https://github.com/Patrick-Erichsen), [@joshavant](https://github.com/joshavant), [@bdjben](https://github.com/bdjben), and [@Coder-Wangyankun](https://github.com/Coder-Wangyankun). ##### Changes - CLI progress: emit Claude CLI commentary progress events and bridge inter-tool commentary into channel progress without exposing internal protocol scaffolding. ([#89834](openclaw/openclaw#89834), [#90883](openclaw/openclaw#90883)) Thanks [@anagnorisis2peripeteia](https://github.com/anagnorisis2peripeteia). - Observability: allow trusted diagnostics channels to capture tool input/output content, add first-assistant-event traces, and warn on slow initial replies. ([#91256](openclaw/openclaw#91256), [#91568](openclaw/openclaw#91568), [#91583](openclaw/openclaw#91583)) Thanks [@amknight](https://github.com/amknight). - Plugins/ClawHub: dogfood reusable package publishing, let dry runs skip publish approval, allow declared installed trusted hooks, report managed plugin version drift, and warn instead of failing on retired Skill Workshop configuration. ([#91574](openclaw/openclaw#91574), [#91591](openclaw/openclaw#91591), [#90004](openclaw/openclaw#90004), [#90927](openclaw/openclaw#90927), [#90838](openclaw/openclaw#90838)) Thanks [@Patrick-Erichsen](https://github.com/Patrick-Erichsen), [@brokemac79](https://github.com/brokemac79), and [@lonexreb](https://github.com/lonexreb). - Memory/providers: move the local llama.cpp runtime into its provider plugin, batch embeddings across files, persist the agent model catalog cache, and keep QMD JSON search one-shot while filtering stale REM recall previews. ([#91324](openclaw/openclaw#91324), [#89138](openclaw/openclaw#89138), [#90457](openclaw/openclaw#90457), [#91837](openclaw/openclaw#91837), [#91851](openclaw/openclaw#91851)) Thanks [@osolmaz](https://github.com/osolmaz), [@mushuiyu886](https://github.com/mushuiyu886), [@ai-hpc](https://github.com/ai-hpc), and [@TurboTheTurtle](https://github.com/TurboTheTurtle). - Channels/mobile: add the QQBot group mention toggle, improve iPad and iPhone control surfaces, and expose the active connection host in the TUI footer. ([#91423](openclaw/openclaw#91423), [#91557](openclaw/openclaw#91557), [#89909](openclaw/openclaw#89909)) Thanks [@cxyhhhhh](https://github.com/cxyhhhhh), [@Solvely-Colin](https://github.com/Solvely-Colin), and [@baskduf](https://github.com/baskduf). - Performance: prewarm TUI runtime plugins, deduplicate plugin auto-enable fanout, trim dense text-delta snapshots, and reuse prepared startup model metadata. ([#90782](openclaw/openclaw#90782), [#89978](openclaw/openclaw#89978), [#91580](openclaw/openclaw#91580), [#91531](openclaw/openclaw#91531)) Thanks [@RomneyDa](https://github.com/RomneyDa) and [@ai-hpc](https://github.com/ai-hpc). ##### Fixes - Agent/session recovery: drop stale approval follow-ups after session rebind, remove drained reply-queue items by identity, recover stale main and visible replies, preserve Codex context-engine compaction ownership, lower the default compaction timeout to 180 seconds while respecting explicit configuration, and keep provider-failure terminal lifecycle state correct. ([#85679](openclaw/openclaw#85679), [#91450](openclaw/openclaw#91450), [#91566](openclaw/openclaw#91566), [#91840](openclaw/openclaw#91840), [#91590](openclaw/openclaw#91590), [#91361](openclaw/openclaw#91361), [#91895](openclaw/openclaw#91895)) Thanks [@openperf](https://github.com/openperf), [@yetval](https://github.com/yetval), [@joshavant](https://github.com/joshavant), [@wangmiao0668000666](https://github.com/wangmiao0668000666), and [@TurboTheTurtle](https://github.com/TurboTheTurtle). - User-visible content boundaries: suppress Codex/Harmony protocol artifacts, neutralize browser and LanceDB memory media directives, redact transcript images, and preserve native `/compact` replies through source suppression. ([#89151](openclaw/openclaw#89151), [#91422](openclaw/openclaw#91422), [#91425](openclaw/openclaw#91425), [#91529](openclaw/openclaw#91529), [#90212](openclaw/openclaw#90212)) Thanks [@joelnishanth](https://github.com/joelnishanth), [@pgondhi987](https://github.com/pgondhi987), [@joshavant](https://github.com/joshavant), and [@snowzlm](https://github.com/snowzlm). - Channel delivery: keep WhatsApp captured replies attached to the successor controller after restart, retry Feishu rate limits, preserve Mattermost thread replies, canonicalize LINE webhook paths, restore Discord reply hydration and runtime timeout exports, and show OpenAI Realtime WebRTC assistant transcripts. ([#85823](openclaw/openclaw#85823), [#89659](openclaw/openclaw#89659), [#91684](openclaw/openclaw#91684), [#91649](openclaw/openclaw#91649), [#90263](openclaw/openclaw#90263), [#91686](openclaw/openclaw#91686), [#90426](openclaw/openclaw#90426)) Thanks [@itsuzef](https://github.com/itsuzef), [@ladygege](https://github.com/ladygege), [@jacobtomlinson](https://github.com/jacobtomlinson), [@fuller-stack-dev](https://github.com/fuller-stack-dev), and [@shushushv](https://github.com/shushushv). - Cron: cancel active task runs cleanly, preserve terminal timeout/cancel state, and recover no-deliver tool warnings instead of silently losing the outcome. ([#90666](openclaw/openclaw#90666), [#90678](openclaw/openclaw#90678)) Thanks [@ai-hpc](https://github.com/ai-hpc). - Gateway/config/auth: share the approval runtime socket token, replace arrays explicitly in `config.patch`, skip the deleted-agent guard only for valid ACP harness sessions, surface headless LaunchAgent state, verify SQLite auth migration before cleanup, and arm QMD startup maintenance. ([#87105](openclaw/openclaw#87105), [#91551](openclaw/openclaw#91551), [#91219](openclaw/openclaw#91219), [#91614](openclaw/openclaw#91614), [#91740](openclaw/openclaw#91740), [#91978](openclaw/openclaw#91978)) Thanks [@fuller-stack-dev](https://github.com/fuller-stack-dev) and [@scotthuang](https://github.com/scotthuang). - Providers/Codex: clarify quota errors, restore the Codex synthetic usage line, canonicalize Codex protocol assets, require API-key auth for realtime voice, normalize ACP model refs, preserve Gemma 4 `reasoning_content`, and avoid guardian review for local models. ([#91390](openclaw/openclaw#91390), [#91709](openclaw/openclaw#91709), [#91507](openclaw/openclaw#91507), [#91567](openclaw/openclaw#91567), [#88630](openclaw/openclaw#88630), [#91696](openclaw/openclaw#91696)) Thanks [@hxy91819](https://github.com/hxy91819), [@brokemac79](https://github.com/brokemac79), [@RomneyDa](https://github.com/RomneyDa), [@joshavant](https://github.com/joshavant), and [@Coder-Wangyankun](https://github.com/Coder-Wangyankun). - Updates/builds: recover package Gateway restarts after refresh failure, expose plugin convergence repair, fall back to Corepack in PATH-less pnpm environments, seed the correct Docker store packages, and keep ClawHub dry-run and publish paths reusable. ([#91581](openclaw/openclaw#91581), [#91599](openclaw/openclaw#91599), [#91547](openclaw/openclaw#91547), [#91591](openclaw/openclaw#91591)) Thanks [@fuller-stack-dev](https://github.com/fuller-stack-dev), [@sallyom](https://github.com/sallyom), and [@Patrick-Erichsen](https://github.com/Patrick-Erichsen). - UI: require explicit user intent before opening chat sessions and drain restored chat queues after session switches. ([#91480](openclaw/openclaw#91480)) Thanks [@TurboTheTurtle](https://github.com/TurboTheTurtle). - Android: avoid the `dataSync` foreground-service type for persistent nodes. ([#80082](openclaw/openclaw#80082)) Thanks [@davelutztx](https://github.com/davelutztx). - Native hooks: bound relay lifetimes so abandoned native hook connections cannot linger indefinitely. ([#91550](openclaw/openclaw#91550)) Thanks [@joshavant](https://github.com/joshavant). </details> --- ### Configuration 📅 **Schedule**: Branch creation - At any time (no schedule defined), Automerge - At any time (no schedule defined). 🚦 **Automerge**: Disabled by config. Please merge this manually once you are satisfied. ♻ **Rebasing**: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox. 🔕 **Ignore**: Close this PR and you won't be reminded about these updates again. --- - [ ] If you want to rebase/retry this PR, check this box --- This PR has been generated by [Renovate Bot](https://github.com/renovatebot/renovate).  Reviewed-on: https://git.erwanleboucher.dev/eleboucher/homelab/pulls/1040

fix(agents): keep context-engine compaction primary

c7666c7

openclaw-barnacle Bot added agents Agent runtime and tooling size: S maintainer Maintainer-authored PR labels Jun 9, 2026

fix(codex): request native compaction after context engines

2e24cea

openclaw-barnacle Bot added extensions: codex size: M and removed size: S labels Jun 9, 2026

joshavant added 2 commits June 9, 2026 00:50

test: cover 90496 compaction and reset edge cases

b5db02f

fix(codex): guard secondary native compaction binding

fc4c870

openclaw-barnacle Bot added size: XL and removed size: M labels Jun 9, 2026

joshavant added 2 commits June 9, 2026 20:08

fix(codex): keep native compaction hint internal

ce92536

fix(codex): wait for active native turns before resume

a7d2bd2

joshavant merged commit a8d33f2 into main Jun 10, 2026
161 checks passed

joshavant deleted the fix/reset-stale-model-provenance-90496 branch June 10, 2026 02:33

github-actions Bot mentioned this pull request Jun 10, 2026

📡 Upstream Digest — 2026-06-10 02:39 UTC curtismercier/openclaw-mods#1050

Open

github-actions Bot mentioned this pull request Jun 10, 2026

🦞 OpenClaw 生态日报 2026-06-10 litang9/big_model_radar#42

Open

jalehman mentioned this pull request Jun 10, 2026

Revisit Codex compaction precedence when a context engine is configured Martian-Engineering/lossless-claw#850

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Fix context-engine compaction ownership for Codex sessions#91590

Fix context-engine compaction ownership for Codex sessions#91590
joshavant merged 6 commits into
mainfrom
fix/reset-stale-model-provenance-90496

joshavant commented Jun 9, 2026 •

edited

Loading

Uh oh!

clawsweeper Bot commented Jun 9, 2026 •

edited

Loading

Uh oh!

joshavant commented Jun 9, 2026

Uh oh!

clawsweeper Bot commented Jun 9, 2026 •

edited

Loading

Uh oh!

joshavant commented Jun 9, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Uh oh!

Conversation

joshavant commented Jun 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Real behavior proof

Added Issue-Shape Regression Proof

Live Discord proof

Regression proof

Verification

Uh oh!

clawsweeper Bot commented Jun 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

joshavant commented Jun 9, 2026

Uh oh!

clawsweeper Bot commented Jun 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

joshavant commented Jun 9, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

joshavant commented Jun 9, 2026 •

edited

Loading

clawsweeper Bot commented Jun 9, 2026 •

edited

Loading

clawsweeper Bot commented Jun 9, 2026 •

edited

Loading