Skip to content

feat(channels): support multi-message sending via split marker#2008

Merged
alexhoshina merged 8 commits intosipeed:mainfrom
uiYzzi:feat/multi-message-sending
Mar 25, 2026
Merged

feat(channels): support multi-message sending via split marker#2008
alexhoshina merged 8 commits intosipeed:mainfrom
uiYzzi:feat/multi-message-sending

Conversation

@uiYzzi
Copy link
Contributor

@uiYzzi uiYzzi commented Mar 25, 2026

📝 Description

Support LLM to proactively send multiple messages via the <|[SPLIT]|> split marker. When SplitOnMarker is enabled in config, the LLM can use this marker in responses to split a single message into multiple independent outbound messages.

Key changes:

  • New pkg/channels/marker.go handles split marker logic
  • Config field Agents.Defaults.SplitOnMarker controls the toggle, disabled by default
  • Frontend config page adds corresponding toggle, labeled "Multi-Message Sending"
  • Message chunks exceeding MaxMessageLength will be further split by length

🗣️ Type of Change

  • 🐞 Bug fix (non-breaking change which fixes an issue)
  • ✨ New feature (non-breaking change which adds functionality)
  • 📖 Documentation update
  • ⚡ Code refactoring (no functional changes, no api changes)

🤖 AI Code Generation

  • 🤖 Fully AI-generated (100% AI, 0% Human)
  • 🛠️ Mostly AI-generated (AI draft, Human verified/modified)
  • 👨‍💻 Mostly Human-written (Human lead, AI assisted or none)

🔗 Related Issue

📚 Technical Context (Skip for Docs)

  • Reference URL:
  • Reasoning: Split marker <|[SPLIT]|> is hardcoded in marker.go, not configurable; the toggle is at AgentDefaults.SplitOnMarker level

☑️ Checklist

  • My code/docs follow the style of this project.
  • I have performed a self-review of my own changes.
  • I have updated the documentation accordingly.

Tests that SplitByMarker and SplitMessage work together correctly, and
that code block boundaries are preserved during marker splitting.
Copy link
Collaborator

@yinwm yinwm left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review: Avoid goto in manager.go

The use of goto lengthSplit in the message processing logic is unnecessary and reduces readability. Go community best practices recommend avoiding goto except for specific cases like unified error handling.

Current approach:

  • Uses goto to jump to fallback logic when marker is not found or disabled
  • Creates non-linear control flow that's harder to follow

Suggested refactoring:
Let the marker splitting try first, then naturally fallback based on result:

var chunks []string

// Step 1: Try marker-based splitting if enabled
if m.config != nil && m.config.Agents.Defaults.SplitOnMarker {
    if markerChunks := SplitByMarker(msg.Content); len(markerChunks) > 1 {
        for _, chunk := range markerChunks {
            chunks = append(chunks, splitByLength(chunk, maxLen)...)
        }
    }
}

// Step 2: Fallback to length-based splitting if no chunks from marker
if len(chunks) == 0 {
    chunks = splitByLength(msg.Content, maxLen)
}

Helper function:

func splitByLength(content string, maxLen int) []string {
    if maxLen > 0 && len([]rune(content)) > maxLen {
        return SplitMessage(content, maxLen)
    }
    return []string{content}
}

Benefits:

  • Linear, easy-to-follow control flow
  • Less nesting, fewer lines
  • splitByLength is independently testable

This is a minor issue but worth addressing for code maintainability.

@alexhoshina alexhoshina merged commit ed618e1 into sipeed:main Mar 25, 2026
4 checks passed
samueltuyizere added a commit to samueltuyizere/picoclaw that referenced this pull request Mar 25, 2026
* update security migration documents

* feat(config): add command pattern detection tool in exec settings (sipeed#1971)

* Add command pattern testing endpoint and UI tool

Adds a new API endpoint `/api/config/test-command-patterns` that tests a
command against configured whitelist and blacklist patterns, along with
a frontend UI component to interactively test patterns.

* Only process deny patterns when enableDenyPatterns is true

* feat(models): add extra_body config field in model add/edit UI (sipeed#1969)

* Add extraBody field to model configuration forms

This adds a new field allowing users to specify additional JSON fields
to inject into the request body when configuring models.

* Handle ExtraBody clearing when frontend sends empty object

The backend now interprets an empty object sent from the frontend as a
signal to clear the ExtraBody field, while nil/undefined preserves the
existing value. Frontend changed to send {} instead of undefined when
the field is empty.

* refactor(web): clean up systray platform build files

Separate embedded tray icons into platform-specific files, rename the
no-cgo systray stub for consistency, and add the app version to the
launcher startup log.

* fix(agent): suppress heartbeat tool feedback (sipeed#1937)

* fix(lint): remove CGO_ENABLED=0 for lint and fix (sipeed#1989)

* fix(lint): remove CGO_ENABLED=0 for lint and fix

* fix makefile

* config: add baidu_search example to config.example.json (sipeed#1990)

Add Baidu Qianfan AI Search configuration block after glm_search,
matching the BaiduSearchConfig struct defaults (enabled: false,
max_results: 10).

Co-authored-by: BeaconCat <BeaconCat@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

* Fix security config precedence during migration (sipeed#1984)

* Fix security config precedence during migration

* add doc

* fix ci

* add baidu search

* feat(web): add WeCom QR binding flow to channel settings (sipeed#1994)

- add backend WeCom QR flow endpoints and in-memory flow state management
- add frontend WeCom binding UI with QR polling and channel enable toggle
- update channel config behavior and i18n strings for WeCom and WeChat
- apply minor formatting cleanup in model-related components

* chore(tui): add build target for picoclaw-launcher TUI and create README for TUI launcher (sipeed#1995)

* fix(build): disable Matrix gateway import on freebsd/arm

Exclude the Matrix gateway shim from freebsd/arm builds because
modernc.org/libc currently fails to compile on that target.
Document the upstream 32-bit FreeBSD codegen mismatch as well.

* fix(release): ignore nightly tags in goreleaser changelog (sipeed#1999)

GoReleaser was picking nightly tags as the "previous tag" when
generating changelogs, causing release changelogs to be incomplete.
Add git.ignore_tags to skip nightly tags.

* feat(tools): add exec tool enhancement with background execution and PTY support (sipeed#1752)

- Unified exec tool with actions: run/list/poll/read/write/send-keys/kill
- PTY support using creack/pty library
- Process session management with background execution
- Process group kill for cleaning up child processes
- Session cleanup: 30-minute TTL for old sessions
- Output buffer: 100MB limit with truncation

Actions:
- run: execute command (sync or background)
- list: list all sessions
- poll: check session status
- read: read session output
- write: send input to session stdin
- send-keys: send special keys (up, down, ctrl-c, enter, etc.)
- kill: terminate session

Tests:
- PTY: allowed commands, write/read, poll, kill, process group kill
- Non-PTY: background execution, list, read, write, poll, kill, process group kill
- Session management: add/get/remove/list/cleanup

* docs: update WeChat community QR code (sipeed#2003)

Co-authored-by: BeaconCat <BeaconCat@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

* feat(logger): add PICOCLAW_LOG_FILE env var for file-only logging

* Feature/add mimo provider (sipeed#1987)

* feat: add Xiaomi MiMo provider support

- Add 'mimo' protocol prefix support in factory_provider.go
- Add default API base URL for MiMo: https://api.xiaomimimo.com/v1
- Update provider-label.ts to include Xiaomi MiMo label
- Add MiMo to provider tables in both English and Chinese documentation
- Add comprehensive unit tests for MiMo provider

MiMo API is compatible with OpenAI API format, making it easy to integrate
with the existing HTTPProvider infrastructure.

Users can now use MiMo by configuring:
{
  "model_name": "mimo",
  "model": "mimo/mimo-v2-pro",
  "api_key": "your-mimo-api-key"
}

* hassas dosyaları kaldırma

* Add .security.yml and onboard to .gitignore

* build(deps): upgrade pty and reorganize sqlite dependencies (sipeed#2012)

- Upgrade github.com/creack/pty from v1.1.9 to v1.1.24
- Move github.com/mattn/go-sqlite3 to indirect dependency
- Move rsc.io/qr from indirect to direct dependency

* feat(channels): support multi-message sending via split marker (sipeed#2008)

* Add multi-message sending via split marker

* Add marker and length split integration tests

Tests that SplitByMarker and SplitMessage work together correctly, and
that code block boundaries are preserved during marker splitting.

* Simplify message chunking logic in channel worker

Extract splitByLength helper function and remove goto-based control
flow.
The logic now flows more naturally - try marker splitting first, then
fall
back to length-based splitting.

* Update multi-message output instructions in agent context

* Add split_on_marker to config defaults

* Add split_on_marker config option

* Rename 'Multi-Message Sending' setting to 'Chatty Mode'

* Add SplitOnMarker config option

---------

Co-authored-by: Cytown <cytown@gmail.com>
Co-authored-by: 柚子 <40852301+uiYzzi@users.noreply.github.com>
Co-authored-by: wenjie <meetwenjie@gmail.com>
Co-authored-by: daming大铭 <yinwm@outlook.com>
Co-authored-by: xiwuqi <64734786+xiwuqi@users.noreply.github.com>
Co-authored-by: taorye <taorye@outlook.com>
Co-authored-by: Luo Peng <luopeng.he@gmail.com>
Co-authored-by: BeaconCat <111232138+BeaconCat@users.noreply.github.com>
Co-authored-by: BeaconCat <BeaconCat@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: lxowalle <83055338+lxowalle@users.noreply.github.com>
Co-authored-by: Guoguo <16666742+imguoguo@users.noreply.github.com>
Co-authored-by: Liu Yuan <namei.unix@gmail.com>
Co-authored-by: 肆月 <2835601846@qq.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants