desktop-v0.0.2

Full Changelog: desktop-v0.0.1...desktop-v0.0.2

desktop-v0.0.1

What's Changed

• fix(web): reload events on SSE reconnect to fill disconnect gap by @howold-lab in #47
• fix(im/discord): resolve three issues preventing Discord bridge from working by @KQAR in #48
• Pr55 history integration by @ChesterRa in #56
• windows适配支持 by @kinghy949 in #55
• Topic/current context prompt pass by @ChesterRa in #57
• fix(encoding):convert files from UTF-8 to UTF-8 BOM. by @yuansui486 in #58

New Contributors

• @KQAR made their first contribution in #48
• @ChesterRa made their first contribution in #56
• @kinghy949 made their first contribution in #55

Full Changelog: v0.4.2...desktop-v0.0.1

v0.4.2

CCCC v0.4.2 Release Notes

v0.4.2 advances the IM integration and operator-facing control surfaces.
It ships key-based IM chat authorization, a reusable actor profiles system, and a remote access control plane — all with matching Web/MCP/CLI coverage.

Highlights

1) Key-based IM chat authorization

IM bridges now authenticate chats through a cryptographic bind-key flow instead of static trust.
Operators generate a one-time bind key in the Web UI, share it with the target IM chat, and the bridge verifies identity before granting access.

Key capabilities:

• /bind <key> command in supported IM platforms (Telegram first)
• automatic subscription on successful bind (no separate /subscribe step)
• pending approval management: operators can review and approve/deny bind requests from the Web Settings → IM Bridge tab
• revoke semantics: authorized chats can be revoked at any time, with immediate effect on message routing
• cccc_im_bind MCP tool for programmatic binding from automation workflows

2) Actor profiles

A new actor profile abstraction lets operators attach reusable identity metadata (display name, avatar, prompt preamble, model preferences) to actors.

Profiles are linked at actor creation or update and propagate across:

• daemon runtime (used for display in IM bridges and Web)
• Web UI (dedicated Actor Profiles settings tab for CRUD operations)
• MCP surface (profile query and assignment tools)

3) Remote access control plane

A new remote access layer allows authorized clients to interact with a CCCC daemon across network boundaries.
Combined with hardened IM revoke semantics, this enables secure multi-node orchestration without exposing the daemon socket directly.

4) Telegram UX improvements

• Typing indicators: the bridge now sends typing actions with configurable throttling, giving collaborators visual feedback during agent processing.
• Actor titles over IDs: IM-rendered messages now prefer human-readable actor titles instead of raw actor identifiers.

5) MCP and operational fixes

• Recipient semantics fix: None (broadcast) vs [] (no recipients) distinction is now correctly preserved in MCP message send, preventing unintended broadcast or silent drops.
• IM KeyManager reload: authorization state is reloaded from disk on each inbound poll, eliminating stale-cache authorization failures after daemon restarts.
• Timestamp handling: authorized_at timestamps in the IM Web UI are now correctly parsed and displayed.

v0.4.1

Highlights

1) Lifecycle correctness and event consistency

Actor/group lifecycle handling was hardened so state transitions and emitted events stay aligned.
This reduces drift between runtime truth and observer surfaces (Web/MCP/stream consumers), especially around start/stop/restart paths.

2) Automation semantics tightened around idle operation

Automation behavior under idle state was clarified and aligned with explicit group-state semantics.
The result is more predictable reminder behavior and fewer operator surprises in long-running groups.

3) Safer secret operations in Web flows

Web actor editing now shows masked secret previews for configured keys.
Operators can verify that the intended key/value shape exists without any plaintext exposure.

Security posture remains unchanged:

• secret values are never returned as plaintext
• write-only private env storage is preserved
• masked previews are metadata-only UX

4) Terminal/runtime guardrails

Terminal maintenance and resize-related handling received additional hardening.
This reduces fragile behavior under mixed runtime conditions and lowers regression risk for active sessions.

v0.4.0

CCCC v0.4.0 Release Notes

A lightweight multi-agent framework with infrastructure-grade reliability.

Chat-native, prompt-driven, and bi-directional by design.

v0.4.0 is the first stable release of the new CCCC generation.
Compared to v0.3 (the tmux-first line in cccc-tmux), this is a ground-up redesign focused on real production workflows, not terminal tricks.

What Actually Changed in v0.4.0

1) Chat-native orchestration: assign work like talking to teammates

v0.4 introduces a full Web control plane where multi-agent coordination happens in a chat flow:

• assign tasks in the conversation, not through fragmented shell rituals
• track delivery/read/ack/reply status as workflow state
• manage agents, group state, and automation from the same workspace

This is not just a UI layer.
It changes the operating model from “operator drives terminals” to “operator drives collaboration state.”

In addition, CCCC supports external IM bridges (Telegram, Slack, Discord, Feishu/Lark, DingTalk), so the same collaboration model can extend beyond the browser when teams are on the move.

2) Workflow-by-design: prompt-configured multi-agent execution

v0.4 treats workflow design as a first-class capability:

• guidance prompts define how agents collaborate, escalate, and report
• automation rules handle recurring, one-time, and interval triggers
• typed actions (notify, group_state, actor_control) make behavior explicit and composable

The key shift is flexibility with control:
teams can build very different multi-agent operating styles without forking the core system.

3) Bi-directional orchestration: CCCC schedules agents, agents can schedule CCCC

v0.4 establishes a two-way control architecture:

• CCCC orchestrates runtimes, routing, delivery obligations, and lifecycle
• agents can actively manage workflows through MCP tools (messaging, automation, group actions)

This unlocks a new class of workflows:
agents are no longer passive executors; they can become active operators that create, tune, and run automation loops safely inside system boundaries.

Why This Is a Major Version Shift (Not a Feature Bump)

Daemon-first system core

All ports (Web/CLI/MCP/IM) are now thin interfaces over one daemon authority.
This removes split-brain behavior and keeps cross-interface semantics consistent.

Append-only ledger as source of truth

Each group persists an auditable event history (groups/<group_id>/ledger.jsonl), enabling replay, incident review, and reliable recovery.

Accountable messaging semantics

Read cursors, acknowledgements, and reply-required obligations make communication operationally enforceable, not best-effort chat noise.

Practical Impact for Teams

With v0.4, teams can run long-lived multi-agent operations with:

• clearer task assignment and follow-up
• stronger delivery accountability
• lower coordination overhead in daily execution
• higher confidence in recovery and debugging

Migration Notes from v0.3

Treat v0.4 as a new generation:

1. Clean old v0.3 runtime residue.
2. Re-initialize groups under the v0.4 model (CCCC_HOME, default ~/.cccc/).
3. Reconnect runtimes via cccc setup --runtime <runtime>.
4. Run cccc doctor before production rollout.

Closing

v0.4.0 turns CCCC into a complete multi-agent collaboration system:

• chat-native control
• prompt-driven workflow design
• bi-directional orchestration between platform and agents

This is the release where CCCC moves from “interesting tooling” to “operable infrastructure.”

v0.3.28

0.3.28 Release Notes

• Added first-class documentation for the new ccontext-based context/ system (replacing legacy POR/SubPOR).
• Recommended setup: install and configure the ccontext MCP for tool-driven context updates: https://github.com/ChesterRa/ccontext (CCCC still works without MCP via the same context/ file mechanism maintained by peers).
• Unified IM context command across bridges: Telegram /context and Slack/Discord !context, with subcommands now|sketch|milestones|tasks|notes|refs|presence.
• Improved IM context rendering: clearer “now” snapshot (Vision / active milestone / active tasks) and more structured outputs for tasks, milestones, notes, refs, and presence.
• Refined system prompts around ccontext hierarchy and ownership rules: PeerA exclusively creates milestones/tasks (to avoid duplicates); PeerA/PeerB both update content.

v0.3.27

Release Notes v0.3.27

🚀 Highlights

Cross-Platform Clipboard Support

We've enhanced the clipboard integration introduced in PR #4 to work seamlessly across all platforms:

• macOS: Uses pbcopy
• WSL2: Automatically detects and uses clip.exe (Windows clipboard bridge)
• Linux Wayland: Uses wl-copy
• Linux X11: Uses xclip or xsel
• Windows: Uses PowerShell clipboard commands

Both TUI mouse selection and tmux copy-mode now auto-detect the appropriate clipboard tool at runtime, ensuring reliable copy functionality regardless of your environment.

Improved TUI Mouse Interaction

• Click-to-Focus: Input field now properly responds to mouse clicks (no longer requires keyboard input to focus)
• Better Event Handling: Fixed mouse event propagation to allow natural interaction with all UI elements
• Cleaner Copy Feedback: Replaced intrusive timeline messages with elegant temporary footer notifications
  • Shows ✓ Copied N lines in the footer for 1.5 seconds
  • Auto-dismisses without polluting message history
  • Non-disruptive to workflow

🎉 Special Thanks

A huge thank you to @waterbang for contributing PR #4, which added mouse drag selection and clipboard integration to the TUI! This release builds upon that foundation with cross-platform enhancements and UX refinements.

v0.3.26

Release Notes v0.3.26

🚀 Highlights

Keepalive Mechanism Overhaul: "Next" Trigger

We've fundamentally redesigned the keepalive trigger logic to be more intent-driven and less intrusive.

• Change: Keepalive now triggers on Next: declarations instead of Progress: markers.
• Rationale: A Next: line represents a commitment to action. If a peer declares intent but stalls, a reminder is warranted. Progress: was often just a status report, leading to unnecessary interruptions when work was actually completed.
• Benefit: Reduced false positives and "nagging", creating a smoother, more natural collaboration rhythm.

Enhanced POR Ownership & Reliability

• PeerA as POR Guardian: PeerA is now explicitly responsible for maintaining the Plan of Record (POR) and Task definitions in both single-peer and dual-peer modes.
• Fix: Resolved a critical bug where POR refresh requests were only sent to PeerB, causing the feature to fail silently in single-peer (PeerA-only) mode.
• Consistency: System refresh cycles now reliably trigger POR updates regardless of the operating mode.

TUI & Terminal Stability

• Robust Shutdown: Added explicit ANSI terminal reset sequences on exit to prevent "corrupted terminal" states (invisible cursor, wrong buffer) after stopping the TUI.
• Shell Initialization: Tmux panes now spawn with bash -ilc (interactive login shell), ensuring user environment variables (PATH, nvm, pyenv) are correctly loaded.

🛠️ Fixes & Improvements

• Documentation: Updated prompt_weaver to reflect the new Next: keepalive trigger in system prompts.
• Code Cleanup: Removed obsolete Progress event parsing logic from keepalive module.
• Refactor: Simplified keepalive context passing in orchestrator.

v0.3.25

v0.3.25

Features:

• Task steps now support 'outputs' field for completed step deliverables
• Task steps now support 'progress' field for in-progress step notes
• TaskManager tracks YAML parse errors with clear actionable messages
• TUI task panel shows parse errors inline with visual indicators
• IM /task command supports task_id parameter (/task T001 or /task 1)
• Handoff injects task parse errors to PeerA for immediate attention

Improvements:

• Status display distinguishes single-peer vs dual-peer mode
• Error messages truncated sensibly for TUI/IM display
• No stderr pollution from parse warnings (errors tracked internally)
• Comprehensive summary view in IM shows all steps per task

v0.3.24

Release Notes - v0.3.24

🎯 Task Mechanism Robustness

This release focuses on making the Blueprint Task mechanism more resilient and practical for real-world agent collaboration.

Robust YAML Parsing

 - Multi-format tolerance: Steps can now be specified as list or dictionary format
 - Flexible status values: Accepts variations like in-progress, in_progress, inprogress, active, wip
 - Step ID normalization: Automatically converts S01, Step1, step-1, s1 → standard S1 format
 - Graceful fallback: Invalid status values default to planned with a warning instead of failing

Role-Specific Task Protocol

 - PeerA (Dual-peer mode): Clear ownership - creates and manages task files
 - PeerB (Dual-peer mode): Collaboration role - can update steps/status but doesn't create tasks
 - Single-peer mode: Full autonomy with combined responsibilities
 - Streamlined prompts: Reduced from 50+ lines to ~25 lines per role for clarity

Planning-First Emphasis

 - Added explicit task check as Step 0 in the one-round loop
 - Removed pending_review status to eliminate workflow bottlenecks
 - Simplified to 4 core statuses: planned, in_progress, blocked, done