#21466: Copilot/fix forwarded message visibility

## Summary Describe the problem and fix in 2–5 bullets: - **Problem:** Forwarded Discord message content never reached the agent; `BodyForAgent` used `baseText` (excludes forwarded content) instead of `messageText` (includes forwarded content). - **Why it matters:** Agent responses to forwarded messages were meaningless — the agent had no visibility into what was actually forwarded. - **What changed:** Three field assignments in `finalizeInboundContext` in `src/discord/monitor/message-handler.process.ts`: ```typescript // Before BodyForAgent: baseText ?? text, RawBody: baseText, CommandBody: baseText, // After BodyForAgent: text, RawBody: text, CommandBody: text, ``` `text` is already assigned `messageText` (from `resolveDiscordMessageText(message, { includeForwarded: true })`). - **What did NOT change:** `Body` (the combined/display body), forwarded media resolution, command authorization, routing, or any other path. ## Change Type (select all) - [x] Bug fix - [ ] Feature - [ ] Refactor - [ ] Docs - [ ] Security hardening - [ ] Chore/infra ## Scope (select all touched areas) - [ ] Gateway / orchestration - [ ] Skills / tool execution - [ ] Auth / tokens - [ ] Memory / storage - [x] Integrations - [ ] API / contracts - [ ] UI / DX - [ ] CI/CD / infra ## Linked Issue/PR - Closes https://github.com/openclaw/openclaw/issues/21439 ## User-visible / Behavior Changes Agents now receive the full forwarded message text when a Discord user forwards a message. Previously the agent saw an empty or user-typed-only message; now it sees the `[Forwarded message from @User]\n<content>` block that `resolveDiscordForwardedMessagesText` already constructed correctly. ## 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: Any - Runtime/container: Node 22+ - Model/provider: Any - Integration/channel: Discord - Relevant config (redacted): Discord bot with message forwarding enabled ### Steps 1. Forward any Discord message to an OpenClaw-connected Discord channel or DM. 2. Observe the agent's reply. ### Expected - Agent response reflects the forwarded message content. ### Actual (before fix) - Agent response ignores the forwarded content (sees empty/typed-only body). ## Evidence - [ ] Failing test/log before + passing after - [ ] Trace/log snippets - [x] Screenshot/recording - [ ] Perf numbers (if relevant) ## Human Verification (required) - Verified scenarios: Tested forwarding: <img width="554" height="392" alt="image" src="https://github.com/user-attachments/assets/33b97529-fbe6-456d-a0fe-81236a159d45" /> - Edge cases checked: None - What you did **not** verify: Re-building openclaw with the code change. I manually modified the `dist` folder of my openclaw installation ## Compatibility / Migration - Backward compatible? Yes - Config/env changes? No - Migration needed? No ## Failure Recovery (if this breaks) - How to disable/revert this change quickly: Revert the three-line change in `src/discord/monitor/message-handler.process.ts`. - Files/config to restore: `src/discord/monitor/message-handler.process.ts` - Known bad symptoms: Agent suddenly sees forwarded text it shouldn't — unlikely given `resolveDiscordMessageText` already had the correct output. ## Risks and Mitigations - Risk: Commands embedded in forwarded messages might now be interpreted as user commands via `CommandBody`. - Mitigation: Command authorization (`commandAuthorized`) is evaluated independently of this field; `CommandBody` is only the text surface, and existing auth gates remain unchanged. <!-- START COPILOT ORIGINAL PROMPT --> <details> <summary>Original prompt</summary> > ## Bug > > When a user forwards a Discord message, the openclaw agent cannot see the forwarded message content. The agent sees an empty or near-empty message even though the forwarded text is correctly parsed. > > ## Root Cause > > In `src/discord/monitor/message-handler.process.ts`, the `BodyForAgent`, `RawBody`, and `CommandBody` fields in the `finalizeInboundContext` call (around line 518-523) use `baseText` instead of the full `messageText`/`text` that includes forwarded content. > > The preflight phase correctly resolves two text values: > - `baseText` = `resolveDiscordMessageText(message, { includeForwarded: false })` — only the user's own typed content > - `messageText` = `resolveDiscordMessageText(message, { includeForwarded: true })` — includes forwarded message text > > However, `BodyForAgent`, `RawBody`, and `CommandBody` are all set to `baseText`, which **excludes** forwarded content. Since `BodyForAgent` is what the LLM/agent actually receives as the user's message, the agent never sees the forwarded message. > > ## Fix > > In `src/discord/monitor/message-handler.process.ts`, change lines 520-522 from: > > ```typescript > const ctxPayload = finalizeInboundContext({ > Body: combinedBody, > BodyForAgent: baseText ?? text, > InboundHistory: inboundHistory, > RawBody: baseText, > CommandBody: baseText, > ``` > > to: > > ```typescript > const ctxPayload = finalizeInboundContext({ > Body: combinedBody, > BodyForAgent: text, > InboundHistory: inboundHistory, > RawBody: text, > CommandBody: text, > ``` > > Where `text` is already assigned `messageText` (which includes forwarded content) on line 321: > ```typescript > const text = messageText; > ``` > > This ensures the agent sees the full message including the `[Forwarded message from @User]\n<forwarded content>` that `resolveDiscordForwardedMessagesText` already correctly constructs. > </details> <!-- START COPILOT CODING AGENT SUFFIX --> *This pull request was created from Copilot chat.* > <!-- START COPILOT CODING AGENT TIPS --> --- 🔒 GitHub Advanced Security automatically protects Copilot coding agent pull requests. You can protect all pull requests by enabling Advanced Security for your repositories. [Learn more about Advanced Security.](https://gh.io/cca-advanced-security) <!-- greptile_comment --> <h3>Greptile Summary</h3> Changed `BodyForAgent`, `RawBody`, and `CommandBody` to use `text` (which includes forwarded message content) instead of `baseText` (which excludes forwarded content). This ensures agents can see the full forwarded message text that Discord users forward, fixing a bug where agents responded to forwarded messages without seeing the forwarded content. The change is surgical and correct: - `text` is assigned from `messageText` on line 321, which calls `resolveDiscordMessageText(message, { includeForwarded: true })` - `baseText` excludes forwarded content by calling `resolveDiscordMessageText(message, { includeForwarded: false })` - Command authorization is computed in preflight using `baseText` (line 491 in message-handler.preflight.ts), so forwarded commands won't bypass security gates - The change only affects what the agent sees, not the display body or authorization logic <h3>Confidence Score: 5/5</h3> - This PR is safe to merge - it's a targeted 3-line fix that correctly addresses forwarded message visibility without introducing security risks - The change is minimal, well-understood, and correctly fixes the forwarded message bug. Command authorization happens in preflight using `baseText` (which excludes forwarded content), so security gates remain intact. The fix only changes what text the agent receives, which is exactly what's needed. The PR includes manual verification with a screenshot. - No files require special attention _{Last reviewed commit: 53fe8e7} <!-- greptile_other_comments_section --> _{(3/5) Reply to the agent's comments like "Can you suggest a fix for this @greptileai?" or ask follow-up questions!} <!-- /greptile_comment -->