fix(macos): use foundationValue when serializing browser proxy POST body by Effet · Pull Request #41635 · openclaw/openclaw

Effet · 2026-03-10T02:23:53Z

Summary

Problem: MacNodeBrowserProxy.makeRequest passes params.body?.value directly to JSONSerialization.data(withJSONObject:). AnyCodable.init(from:) stores decoded JSON objects as [String: AnyCodable] internally — a pure Swift struct (__SwiftValue) that NSJSONSerialization can't handle, causing an NSInvalidArgumentException crash (SIGABRT) on every non-GET request with a JSON body.
Why it matters: Any POST/PUT/PATCH request routed through the Mac browser proxy crashes the app. The crash is deterministic and not edge-case.
What changed: Use params.body?.foundationValue instead of params.body?.value. foundationValue (already defined in AnyCodable+Helpers.swift) recursively converts [String: AnyCodable] → [String: Any] so JSONSerialization receives only Foundation-compatible types.
What did NOT change: No behavior change for GET requests or response parsing; no new dependencies.

Change Type (select all)

Bug fix

Scope (select all touched areas)

UI / DX

Linked Issue/PR

Closes macOS app crashes on browser.proxy invoke with Invalid type in JSON write (__SwiftValue) #40957

User-visible / Behavior Changes

Mac app no longer crashes when making POST/PUT/PATCH requests through the browser proxy.

Security Impact (required)

New permissions/capabilities? No
Secrets/tokens handling changed? No
New/changed network calls? No
Command/tool execution surface changed? No
Data access scope changed? No

Repro + Verification

Environment

OS: macOS (latest)
Runtime: Mac app (MacNodeRuntime)

Steps

Open Mac app with gateway running
Trigger any browser proxy call that uses a non-GET method with a JSON body (e.g. a POST via MacNodeBrowserProxy.request)
App crashes with NSInvalidArgumentException: Invalid type in JSON write (__SwiftValue)

Expected

Request serializes and sends successfully

Actual

SIGABRT crash in MacNodeBrowserProxy.makeRequest at JSONSerialization.data(withJSONObject:)

Evidence

Regression test added: postRequestSerializesNestedBodyWithoutCrash in MacNodeBrowserProxyTests.swift — covers nested dict + array body, was impossible to add before the fix (would crash)

Human Verification (required)

Verified scenarios: read crash stack trace, identified root cause in source, confirmed foundationValue is the correct converter, added failing→passing test
Edge cases checked: nested objects, arrays, null values all handled by existing foundationValue recursion
What you did not verify: live end-to-end POST request through running gateway (no Mac build in this session)

Review Conversations

I replied to or resolved every bot review conversation I addressed in this PR.
I left unresolved only the conversations that still need reviewer or maintainer judgment.

Compatibility / Migration

Backward compatible? Yes
Config/env changes? No
Migration needed? No

Failure Recovery (if this breaks)

Revert: restore params.body?.value in MacNodeBrowserProxy.swift:149
Known bad symptom: app crashes on any browser proxy POST request (same as before fix)

Risks and Mitigations

Risk: foundationValue on an unrecognized type falls through to self.value (the default case) — same behavior as before, so no regression risk.

greptile-apps · 2026-03-10T02:27:03Z

Greptile Summary

This PR fixes a deterministic SIGABRT crash in MacNodeBrowserProxy.makeRequest when making any non-GET HTTP request with a JSON body. The root cause was passing AnyCodable.value directly to JSONSerialization: since AnyCodable.init(from:) stores decoded JSON objects as [String: AnyCodable] (a pure Swift struct), NSJSONSerialization would throw NSInvalidArgumentException: Invalid type in JSON write (__SwiftValue) on every such call. The fix correctly substitutes body.foundationValue, which already exists in AnyCodable+Helpers.swift and recursively converts [String: AnyCodable] → [String: Any] so only Foundation-compatible types reach the serializer. A targeted regression test is included.

Fix (MacNodeBrowserProxy.swift line 149–150): params.body?.value replaced with params.body + body.foundationValue; semantics of when the body is set are unchanged (both paths skip the body iff params.body is nil)
foundationValue: Handles all five decoded cases from AnyCodable.init(from:) — dicts, arrays, and primitives/null — recursively, with the default branch returning self.value (same as before), so no regression for unrecognized types
Regression test: Covers nested dict + array bodies; previously would crash before the fix; validates round-tripped values after serialization/deserialization

Confidence Score: 4/5

Safe to merge — fixes a deterministic crash with a minimal, well-understood one-line change backed by a meaningful regression test.
This PR is safe to merge. The fix is correct: foundationValue recursively converts [String: AnyCodable] to [String: Any], and the default branch preserves prior behavior for primitives/null. The change in optional binding semantics (params.body?.value → params.body + .foundationValue) is equivalent — both paths skip body attachment iff params.body is nil. The regression test is meaningful and would catch a recurrence. Only live end-to-end POST testing is missing, which the author explicitly acknowledges.
No files require special attention.

_{Last reviewed commit: 991d42b}

@frankekn

…dary marker variants (openclaw#35983) Merged via squash. Prepared head SHA: ff07dc4 Co-authored-by: urianpaul94 <[email protected]> Co-authored-by: frankekn <[email protected]> Reviewed-by: @frankekn

@obviyus

…penclaw#40740) Merged via squash. Prepared head SHA: a4456d4 Co-authored-by: sircrumpet <[email protected]> Co-authored-by: obviyus <[email protected]> Reviewed-by: @obviyus

@frankekn

Merged via squash. Prepared head SHA: 0a8ebf8 Co-authored-by: frankekn <[email protected]> Co-authored-by: frankekn <[email protected]> Reviewed-by: @frankekn

@BunsDev

Merged via squash. Prepared head SHA: 8abc6c1 Co-authored-by: BunsDev <[email protected]> Co-authored-by: BunsDev <[email protected]> Reviewed-by: @BunsDev

@BunsDev

…penclaw#14967) Merged via squash. Prepared head SHA: 1bb49b2 Co-authored-by: rixau <[email protected]> Co-authored-by: BunsDev <[email protected]> Reviewed-by: @BunsDev

@mcaxtr

…ess is ro (openclaw#40757) Merged via squash. Prepared head SHA: 0e8b27b Co-authored-by: dsantoreis <[email protected]> Co-authored-by: mcaxtr <[email protected]> Reviewed-by: @mcaxtr

@mukhtharcm

…claw#41176) Merged via squash. Prepared head SHA: 33cac4c Co-authored-by: hnykda <[email protected]> Co-authored-by: mukhtharcm <[email protected]> Reviewed-by: @mukhtharcm

@BradGroux

… Framework compatibility (openclaw#41838) * fix(msteams): use General channel conversation ID as team key for Bot Framework compatibility Bot Framework sends `activity.channelData.team.id` as the General channel's conversation ID (e.g. `19:[email protected]`), not the Graph API group GUID (e.g. `fa101332-cf00-431b-b0ea-f701a85fde81`). The startup resolver was storing the Graph GUID as the team config key, so runtime matching always failed and every channel message was silently dropped. Fix: always call `listChannelsForTeam` during resolution to find the General channel, then use its conversation ID as the stored `teamId`. When a specific channel is also configured, reuse the same channel list rather than issuing a second API call. Falls back to the Graph GUID if the General channel cannot be found (renamed/deleted edge case). Fixes openclaw#41390 * fix(msteams): handle listChannelsForTeam failure gracefully * fix(msteams): trim General channel ID and guard against empty string * fix: document MS Teams allowlist team-key fix (openclaw#41838) (thanks @BradGroux) --------- Co-authored-by: bradgroux <[email protected]> Co-authored-by: Onur <[email protected]>

@mukhtharcm

Merged via squash. Prepared head SHA: 5cffcb0 Co-authored-by: teconomix <[email protected]> Co-authored-by: mukhtharcm <[email protected]> Reviewed-by: @mukhtharcm

@dutifulbob

* ACPX: bump bundled acpx to 0.1.16 * fix: bump acpx pin to 0.1.16 (openclaw#41975) (thanks @dutifulbob) --------- Co-authored-by: Onur <[email protected]>

@pejmanjohn

…me (openclaw#41847) * feat(acp): add resumeSessionId to sessions_spawn for ACP session resume Thread resumeSessionId through the ACP session spawn pipeline so agents can resume existing sessions (e.g. a prior Codex conversation) instead of starting fresh. Flow: sessions_spawn tool → spawnAcpDirect → initializeSession → ensureSession → acpx --resume-session flag → agent session/load - Add resumeSessionId param to sessions-spawn-tool schema with description so agents can discover and use it - Thread through SpawnAcpParams → AcpInitializeSessionInput → AcpRuntimeEnsureInput → acpx extension runtime - Pass as --resume-session flag to acpx CLI - Error hard (exit 4) on non-existent session, no silent fallback - All new fields optional for backward compatibility Depends on acpx >= 0.1.16 (openclaw/acpx#85, merged, pending release). Tests: 26/26 pass (runtime + tool schema) Verified e2e: Discord → sessions_spawn(resumeSessionId) → Codex resumed session and recalled stored secret. 🤖 AI-assisted * fix: guard resumeSessionId against non-ACP runtime Add early-return error when resumeSessionId is passed without runtime="acp" (mirrors existing streamTo guard). Without this, the parameter is silently ignored and the agent gets a fresh session instead of resuming. Also update schema description to note the runtime=acp requirement. Addresses Greptile review feedback. * ACP: add changelog entry for session resume (openclaw#41847) (thanks @pejmanjohn) --------- Co-authored-by: Pejman Pour-Moezzi <[email protected]> Co-authored-by: Onur <[email protected]>

@mukhtharcm

…claw#18655) Merged via squash. Prepared head SHA: d30fff1 Co-authored-by: echo931 <[email protected]> Co-authored-by: mukhtharcm <[email protected]> Reviewed-by: @mukhtharcm

@altaywtf

…k runs (openclaw#41711) Merged via squash. Prepared head SHA: 8be8967 Co-authored-by: cgdusek <[email protected]> Co-authored-by: altaywtf <[email protected]> Reviewed-by: @altaywtf

@hydro13

Merged via squash. Prepared head SHA: 85c4eb7 Co-authored-by: smysle <[email protected]> Co-authored-by: hydro13 <[email protected]> Reviewed-by: @hydro13

…w#42173) Models like GLM-5 and DeepSeek sometimes emit internal delimiter tokens in their responses. Uses generic pattern in the text extraction pipeline, following the same architecture as stripMinimaxToolCallXml. Closes openclaw#40020 Supersedes openclaw#40573 Co-authored-by: imwyvern <[email protected]>

@imwyvern

…kens (openclaw#42216) Thanks @imwyvern.

…34369) Strip trailing /v1beta from baseUrl before appending the version segment, so callers that already include /v1beta in their base URL (e.g. subagent-registry) no longer produce /v1beta/v1beta/models/… which results in a 404 from the Gemini API. Closes openclaw#34312 Co-authored-by: Claude Opus 4.6 <[email protected]>

@pejmanjohn

* docs(acp): document resumeSessionId for session resume * docs: clarify ACP resumeSessionId thread/mode behavior (openclaw#42280) (thanks @pejmanjohn) --------- Co-authored-by: Onur <[email protected]>

@altaywtf

…penclaw#42278) Merged via squash. Prepared head SHA: f3cdfa7 Co-authored-by: CryUshio <[email protected]> Co-authored-by: altaywtf <[email protected]> Reviewed-by: @altaywtf

openclaw-barnacle · 2026-03-11T09:08:23Z