fix(reply): keep webchat-origin main session replies on originating surface by bmendonca3 · Pull Request #33250 · openclaw/openclaw

bmendonca3 · 2026-03-03T16:24:33Z

Summary

Describe the problem and fix in 2–5 bullets:

Problem: Main-session turns that originate from webchat (TUI/gateway client) can reuse a previously persisted external delivery route (for example WhatsApp), which causes replies to fan out to the wrong surface.
Why it matters: Control-surface messages should stay on the control surface; leaking those replies to external channels is user-visible and can send unintended content.
What changed: Adjusted resolveLastChannelRaw and resolveLastToRaw to prefer webchat origin routing when the session key is the main session (agent:<id>:main), then added a regression test that reproduces the leak and validates webchat pinning.
What did NOT change (scope boundary): Internal non-webchat handoff channels (for example sessions_send) still preserve persisted external fallback behavior; non-main session behavior is unchanged.

Change Type (select all)

Scope (select all touched areas)

Linked Issue/PR

User-visible / Behavior Changes

Main-session replies triggered from webchat/TUI now stay on webchat instead of reusing a stale external route (for example WhatsApp).

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)
If any Yes, explain risk + mitigation:

Repro + Verification

Environment

OS: macOS (dev)
Runtime/container: Node 22 + pnpm
Model/provider: N/A (routing/session behavior)
Integration/channel (if any): webchat main session + persisted external route
Relevant config (redacted): session store with main entry persisted as external channel route

Steps

Seed session store so agent:main:main has persisted lastChannel=whatsapp, lastTo=<phone>.
Initialize a turn on that same session with OriginatingChannel=webchat and OriginatingTo=session:webchat-main.
Inspect resolved session delivery context.

Expected

Main-session reply route remains on webchat for that turn (lastChannel=webchat, lastTo=session:webchat-main).

Actual

Before fix, routing reused persisted external route (lastChannel=whatsapp) and could deliver to the wrong channel.

Evidence

Attach at least one:

Failing test/log before + passing after
Trace/log snippets
Screenshot/recording
Perf numbers (if relevant)

Human Verification (required)

What you personally verified (not just CI), and how:

Verified scenarios:
- Added regression test: prefers webchat route over persisted external route for main session turns.
- Ran full src/auto-reply/reply/session.test.ts to ensure routing behavior remains coherent.
Edge cases checked:
- Non-webchat internal channels still preserve persisted external fallback behavior.
- Non-main session behavior remains unchanged.
What you did not verify:
- Live multi-channel delivery against real WhatsApp/TUI runtime.

Compatibility / Migration

Backward compatible? (Yes)
Config/env changes? (No)
Migration needed? (No)
If yes, exact upgrade steps:

Failure Recovery (if this breaks)

How to disable/revert this change quickly: Revert this PR commit.
Files/config to restore: src/auto-reply/reply/session-delivery.ts, src/auto-reply/reply/session.test.ts.
Known bad symptoms reviewers should watch for: webchat main-session turns unexpectedly delivering to stale external channels.

Risks and Mitigations

List only real risks for this PR. Add/remove entries as needed. If none, write None.

Risk: Main-session webchat messages no longer inherit a previous external route in cases where users intentionally relied on that implicit fallback.
- Mitigation: Scope is intentionally narrow to main-session + webchat origin only; tests preserve existing behavior for non-webchat internal channels.

greptile-apps · 2026-03-03T16:28:46Z

Greptile Summary

This PR fixes a routing leak where main-session turns originating from webchat could reuse a previously persisted external delivery route (e.g., WhatsApp), causing replies to fan out to the wrong surface. The fix adds narrow early-return guards in resolveLastChannelRaw and resolveLastToRaw that short-circuit the external-fallback logic whenever the originating channel is webchat and the session key identifies the main session.

Key points:

Channel guard (resolveLastChannelRaw): Correctly returns originatingChannelRaw ("webchat") before the external-fallback branch, preventing stale external channels from overwriting it.
Potential to mismatch: The resolveLastToRaw early-return guard still falls back to persistedLastTo when no current destination is provided. In edge cases where a webchat main-session turn arrives without OriginatingTo or To but has persisted external lastTo data, this creates a logical mismatch: lastChannel="webchat" paired with an external address. The new regression test covers the primary scenario (where OriginatingTo is provided), but not this edge case.
Backward compatibility: Non-webchat internal channels and non-main sessions are unaffected.

Confidence Score: 3/5

Safe for the documented main-session webchat routing scenario, but carries an unaddressed channel/to mismatch risk for edge cases without OriginatingTo.
The primary bug (stale external channel leak on webchat main-session turns) is correctly fixed and covered by the regression test when OriginatingTo is provided. However, a secondary logic error exists in resolveLastToRaw: when a webchat main-session turn omits both OriginatingTo and To but inherits persisted external lastTo data, the code produces an inconsistent lastChannel="webchat" / lastTo=<external> pairing. This edge case is demonstrable via the existing test "keeps webchat channel for webchat/main sessions" (which passes no OriginatingTo) combined with persisted external data. Score is lowered from 4 to 3 because this mismatch represents a real logic error that could cause delivery failures, even though it is narrower in scope than the original leak.
src/auto-reply/reply/session-delivery.ts — resolveLastToRaw function should guard or remove the persistedLastTo fallback in the early-return branch for webchat main sessions.

_{Last reviewed commit: c4d04bc}

greptile-apps

_{2 files reviewed, 2 comments}

_{Edit Code Review Agent Settings | Greptile}

src/auto-reply/reply/session-delivery.ts

bmendonca3 · 2026-03-03T19:43:00Z

@greptileai

src/auto-reply/reply/session-delivery.ts

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 5f9f6c5e07

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

src/auto-reply/reply/session-delivery.ts

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: a993ea6c56

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

src/auto-reply/reply/session-delivery.ts

Takhoffman · 2026-03-04T02:57:19Z

Superseded by #33786.

The synthesis PR includes webchat/main reply-surface routing fixes and aligned session routing invariants.

Takhoffman · 2026-03-04T03:41:16Z

Thanks again for this work.

This PR was triaged and merged using an AI-assisted review/synthesis workflow.

This landed indirectly via the synthesized PR #33786, and your contribution is credited in the changelog and as a co-author on the merge commit.

Closing this PR as superseded by #33786 (reason: superseded). If anything here looks incorrect or incomplete, reply to reopen and we can reassess.

Takhoffman · 2026-03-04T03:41:51Z

Correction for accuracy: the synthesized work is credited in the changelog, but the squash merge commit for #33786 does not include explicit Co-authored-by trailers.

reply: keep main session webchat replies on originating surface

7c13cce

openclaw-barnacle bot added size: XS experienced-contributor labels Mar 3, 2026

greptile-apps bot reviewed Mar 3, 2026

View reviewed changes

src/auto-reply/reply/session-delivery.ts Outdated Show resolved Hide resolved

reply: drop unreachable lastChannel fallback for main sessions

c4d04bc

greptile-apps bot reviewed Mar 3, 2026

View reviewed changes

src/auto-reply/reply/session-delivery.ts Show resolved Hide resolved

reply: avoid stale external lastTo on webchat main turns

5f9f6c5

openclaw-barnacle bot added size: S and removed size: XS labels Mar 3, 2026

chatgpt-codex-connector bot reviewed Mar 3, 2026

View reviewed changes

src/auto-reply/reply/session-delivery.ts Show resolved Hide resolved

gateway: preserve webchat origin in chat send routing

a993ea6

openclaw-barnacle bot added app: web-ui App: web-ui gateway Gateway runtime labels Mar 4, 2026

chatgpt-codex-connector bot reviewed Mar 4, 2026

View reviewed changes

src/auto-reply/reply/session-delivery.ts Show resolved Hide resolved

Takhoffman mentioned this pull request Mar 4, 2026

fix(routing): unify session delivery invariants for duplicate suppression #33786

Merged

Takhoffman closed this Mar 4, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

fix(reply): keep webchat-origin main session replies on originating surface#33250

fix(reply): keep webchat-origin main session replies on originating surface#33250
bmendonca3 wants to merge 4 commits intoopenclaw:mainfrom
bmendonca3:bm/33201-main-session-origin-route

bmendonca3 commented Mar 3, 2026

Uh oh!

greptile-apps bot commented Mar 3, 2026 •

edited

Loading

Uh oh!

greptile-apps bot left a comment •

edited

Loading

Uh oh!

Uh oh!

bmendonca3 commented Mar 3, 2026

Uh oh!

Uh oh!

chatgpt-codex-connector bot left a comment

Uh oh!

Uh oh!

chatgpt-codex-connector bot left a comment

Uh oh!

Uh oh!

Takhoffman commented Mar 4, 2026

Uh oh!

Takhoffman commented Mar 4, 2026

Uh oh!

Takhoffman commented Mar 4, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

bmendonca3 commented Mar 3, 2026

Summary

Change Type (select all)

Scope (select all touched areas)

Linked Issue/PR

User-visible / Behavior Changes

Security Impact (required)

Repro + Verification

Environment

Steps

Expected

Actual

Evidence

Human Verification (required)

Compatibility / Migration

Failure Recovery (if this breaks)

Risks and Mitigations

Uh oh!

greptile-apps bot commented Mar 3, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Greptile Summary

Confidence Score: 3/5

Uh oh!

greptile-apps bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

bmendonca3 commented Mar 3, 2026

Uh oh!

Uh oh!

chatgpt-codex-connector bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

Uh oh!

chatgpt-codex-connector bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

Uh oh!

Takhoffman commented Mar 4, 2026

Uh oh!

Takhoffman commented Mar 4, 2026

Uh oh!

Takhoffman commented Mar 4, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

greptile-apps bot commented Mar 3, 2026 •

edited

Loading

greptile-apps bot left a comment •

edited

Loading