WorkBuddy Session Callback lets one session, cron job, or external process wake another WorkBuddy session and inject a message with full context.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install workbuddy-skill-session-callback
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install workbuddy-skill-session-callback using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is WorkBuddy Session Callback?
WorkBuddy Session Callback is a high-impact workflow skill for the WorkBuddy desktop app that enables one session or trusted local process to trigger another session, inject a user-style message, and continue execution with preserved context.
Openclaw Skills users can use this skill to build reliable cross-session coordination for wakeups, scheduled callbacks, asynchronous notifications, and task handoff patterns without relying on always-on agents or unsupported platform behavior. It is designed specifically for WorkBuddy’s local session registry, ACP endpoint, and queue-aware session handling.
WorkBuddy Session Callback Use Cases
- Wake a dormant WorkBuddy session and continue a task with full conversational context.
- Inject a notification into a specific target session from a trusted cron job or local script.
- Send completion alerts from an asynchronous process back into the owning session.
- Coordinate multiple WorkBuddy sessions for task handoff or staged workflows.
- Replace native
sessions_send-style behavior in WorkBuddy desktop automation flows. - Trigger a session callback when a scheduled automation needs to resume work later.
- Deliver context-preserving reminders or follow-ups to a running agent session.
How WorkBuddy Session Callback Works
- Identify the target session by reading the local WorkBuddy session registry and finding an active endpoint.
- Verify the session is alive and reachable; if the endpoint is unavailable, the session must be reopened before injection.
- Run the callback script or call the ACP endpoint manually to connect, initialize, load the session, and prepare the prompt.
- Inject the message as a user-style prompt so the target agent handles it naturally in its own context.
- Confirm the callback succeeded by checking session response updates and the generated JSONL transcript.
- If the session is busy, let the message queue until the session becomes idle instead of forcing interruption.
- Use WorkBuddy built-in automation for scheduled callbacks, then clean up temporary automation and logs when done.
WorkBuddy Session Callback Setup
- Ensure WorkBuddy desktop is installed and running on Windows or macOS.
- Confirm the local session files exist under
~/.workbuddy/sessions/and the session index exists at~/.workbuddy/app/sessions.json. - Use the included callback script to discover the active endpoint and inject a message.
python scripts/callback.py <session_id> "message to inject" [--workdir <path>]
- For manual integration, use the ACP flow exposed by WorkBuddy’s local endpoint:
POST {endpoint}/api/v1/acp/connect
POST {endpoint}/api/v1/acp
POST {endpoint}/api/v1/acp
POST {endpoint}/api/v1/acp
- Make sure the target session UUID is correct and not confused with a process ID.
- Use localhost-only access; this skill is intended for local loopback operation and should not be exposed to the network.
- For timed callbacks, configure WorkBuddy built-in automation rather than background shell daemons.
# Example: schedule inside WorkBuddy automation, then invoke callback logic locally
python scripts/callback.py <session_uuid> "scheduled reminder or follow-up"
- After use, remove temporary automation rules and clear any local logs created during testing.
WorkBuddy Session Callback Data Schema & Taxonomy
The skill relies on WorkBuddy-local state and transcript files rather than a cloud database.
| Artifact | Location | Purpose |
|---|---|---|
| Session registry | ~/.workbuddy/sessions/ |
Stores live session registration records and endpoints |
| Session index | ~/.workbuddy/app/sessions.json |
Maps sessions to UUIDs and heartbeat/metadata information |
| Transcript logs | projects/*.jsonl |
Records user and assistant turns for session continuation |
| ACP endpoint | /api/v1/acp/* |
Local session control and prompt injection interface |
Metadata taxonomy
sessionId: The target session UUID used for callback routing.endpoint: The local HTTP endpoint for the active session process.lastHeartbeat: Used to sort and detect the freshest registered session.cwd: The working directory associated with the session context.mcpServers: Session environment metadata passed during load.prompt: The injected message content, typically delivered as user text.role=user/role=assistant: Transcript roles stored in JSONL output.
Data behavior
- The skill loads the full transcript history so the target agent keeps context continuity.
- Injected content is queued when the target session is busy.
- The transcript is archived automatically after the callback completes.
WorkBuddy Session Callback Advanced Features
- Cross-session wakeup support for Openclaw Skills-style callback workflows inside WorkBuddy desktop.
- Queue-aware injection: messages are delayed rather than dropped when the target session is busy.
- Full-context continuation by loading the session transcript before prompt injection.
- Built-in automation support for scheduled callbacks without relying on unmanaged shell daemons.
- Local loopback operation for reduced exposure and tighter safety boundaries.
- One-click callback script that discovers the endpoint, connects, initializes, loads, and prompts in sequence.
- Works as a native alternative to external
sessions_send-style mechanisms for WorkBuddy. - Explicit cleanup pattern for temporary automations, logs, and permission relaxations after use.
- Designed for high-impact operations with strict target-session scoping and trusted-source requirements.
SKILL.md
Loading
Related Openclaw Skills
Agentic Ledger
Agentic Ledger is a local-first proxy and analytics layer for OpenClaw that tracks usage, enforces budgets, and replays runs.
China Weather Data
Openclaw Skills for China Weather Data lets you query, download, and automate China meteorological station weather data with an official CMA source and a free Open-Meteo fallback.
GeoTIFF Metadata Viewer
Openclaw Skills for GeoTIFF Metadata Viewer lets you inspect GeoTIFF file metadata, CRS, resolution, tags, and corner coordinates with zero dependencies.
China Administrative Divisions Downloader
Openclaw Skills for downloading and exporting Chinese administrative-division vector data with search, metadata lookup, bbox calculation, and batch GIS output.
yt-agent
yt-agent turns YouTube videos and subscriptions into grounded summaries, highlights, Q&A, presentations, and digests through the yt-mem-ai CLI.
Privacymask — Data De-identification and Privacy Protection Assistant
Privacymask is a local-first data masking skill that detects and redacts sensitive information across common file formats without uploading data.
Featured*
