Agent Telegram standardizes how eight specialized AI agents send routed, recognizable progress updates to a Telegram user.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install telegram-agent-comm
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 telegram-agent-comm 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 Agent Telegram?
Agent Telegram is a communication specification for multi-agent AI teams using Telegram as a reliable user-notification channel. It maps eight agent roles to dedicated account IDs, bot identities, emoji prefixes, and standardized reporting templates so users can quickly understand who sent a message and what happened.
As an Openclaw Skills integration, it requires every outbound message to use the message tool with channel: "telegram", the correct accountId, and the fixed target 5440561025. The workflow covers task acceptance, subtask completion, blockers, and final result aggregation while keeping role-based communication consistent across Windows, macOS, and Linux.
Agent Telegram Use Cases
- Multi-agent software development: Let main coordinate architect, product, backend, frontend, and QA work while each role reports progress independently.
- Automated data collection: Have the crawler agent announce completed collection jobs and provide output file paths.
- Backend and frontend delivery updates: Notify users when APIs, interface work, or technical documentation are complete.
- QA escalation: Report blocking defects with a clear problem description and recommended decision or remediation.
- Task lifecycle notifications: Send standardized messages when work starts, subtasks finish, issues arise, and all work is complete.
- Centralized Telegram routing: Use separate bot accounts for agent identities while delivering notifications to a controlled Telegram target.
- Communication auditability: Use optional callbacks, delivery status handling, and communication archiving where supported by the paid platform tier.
- Not suitable for: Spam or mass broadcasting, non-Telegram delivery, end-to-end encrypted sensitive communications, or large cross-team announcements.
How Agent Telegram Works
- Identify the active role: Determine the current agent's role and select its mapped
accountIdand emoji. If the role is unknown, usedefaultand the 🤖 prefix. - Select the reporting event: Choose the task-start, task-complete, issue, or final-summary message type based on the workflow stage.
- Assemble the message: Include the task name, subtask, output path, issue details, or recommendation, and place the role emoji at the beginning.
- Route through the message tool: Send the notification with
action: "send",channel: "telegram", the role-specificaccountId, andtarget: "5440561025". - Validate routing metadata: Confirm the exact account ID, especially the product role's required
sproductvalue rather thanproduct. - Report continuously: Notify the user when work is received, when subtasks complete, and whenever a blocker needs attention.
- Aggregate results: The main agent collects role updates and sends a consolidated final report to the user.
- Handle failures: Inspect input parameters, logs, credentials, network access, file paths, and rate limits before retrying or escalating.
Agent Telegram Setup
- Install or enable the skill: Place
SKILL.mdat the agent platform's skill workspace path:
mkdir -p "$HOME/.skill-platform/workspace/skills/agent-telegram"
# Copy the Agent Telegram SKILL.md into the directory above
Create Telegram bots: Use Telegram's
@BotFatherto create one bot token for each agent role that needs an independent identity.Configure Telegram accounts: Edit the platform configuration at
$HOME/.skill-platform/skill-platform.jsonand add accounts underchannels.telegram.accounts. Store each bot token in the corresponding account'stokenfield.
mkdir -p "$HOME/.skill-platform"
${EDITOR:-vi} "$HOME/.skill-platform/skill-platform.json"
Register the role mapping: Configure these account IDs:
defaultfor main,architect,backend,frontend,sproductfor product,content,crawler, andqa.Verify platform dependencies: Confirm that the agent platform provides the
messagetool, the LLM is available, and outbound HTTPS access tohttps://api.telegram.orgis permitted.Run a controlled test: Send a short test message using the standard schema, verify the sender identity and emoji, and confirm delivery to Telegram target
5440561025.Secure the deployment: Keep bot tokens out of source control, restrict skill access, use HTTPS, and avoid transmitting sensitive data. Messages are limited to 4096 characters and cannot be recalled or edited after sending.
Agent Telegram Data Schema & Taxonomy
Request schema
| Field | Type | Required | Description |
|---|---|---|---|
input |
string | Yes | Agent Telegram input data or command. |
options |
object | No | Additional settings such as mode or formatting preferences. |
callback_url |
string | No | URL for asynchronous completion notifications. |
Message routing schema
| Field | Required value or format | Purpose |
|---|---|---|
action |
send |
Requests outbound delivery. |
channel |
telegram |
Selects the Telegram transport. |
accountId |
Role-specific ID | Selects the sending bot and agent identity. |
target |
5440561025 |
Fixed user Telegram ID defined by this specification. |
message |
Emoji plus structured text | Contains the user-facing update. |
Agent taxonomy
| Agent role | accountId |
Emoji | Responsibility |
|---|---|---|---|
| main | default |
🤖 | Task distribution and final aggregation |
| architect | architect |
🏗️ | System design and technology selection |
| backend | backend |
🔧 | APIs and server development |
| frontend | frontend |
🎨 | UI and interaction development |
| product | sproduct |
🟡 | Requirements and prioritization |
| content | content |
✍️ | Copy and content assets |
| crawler | crawler |
🕷️ | Data collection |
| qa | qa |
🧪 | Quality verification |
Generated and referenced data
- Configuration:
$HOME/.skill-platform/skill-platform.json, specificallychannels.telegram.accountsand each account'stoken. - Skill definition:
$HOME/.skill-platform/workspace/skills/agent-telegram/SKILL.md. - Output references: Messages may include paths such as
$HOME/Desktop/project/data/output.jsonor documentation files. - Execution result: A structured result includes
success,data.result,execution_time,metadata.version,metadata.processor,execution_log, anderror. - Message templates: Task start, task completion, and issue reports use consistent emoji, labels, output paths, descriptions, and recommendations.
Agent Telegram Advanced Features
- Eight-account role routing: Separate bot tokens and account IDs prevent simultaneous agent messages from interfering with one another.
- Main-agent orchestration: The
mainrole distributes work, receives updates, and publishes a consolidated summary. - Template-driven reporting: Standard task-start, completion, and issue templates improve readability and reduce formatting errors.
- Multi-channel batch delivery: The paid tier supports multi-channel message batching in addition to role-based Telegram delivery.
- Variable injection: Paid features support message templates with dynamic task names, paths, statuses, and result data.
- Delivery callbacks: Optional
callback_urlsupport and paid-tier real-time delivery status callbacks enable asynchronous monitoring. - Communication archiving and retrieval: The paid tier supports storing and searching communication history.
- File-aware notifications: Messages can include generated file paths, while supported deployments may use an
attachmentfield or a separate upload tool for direct attachments. - Operational diagnostics: Troubleshoot authentication failures, rate limits such as 429, timeouts such as 504, missing files, permissions, command failures, and network connectivity issues.
- Cross-platform execution: The Markdown-based skill works with compatible AI agents such as Claude Code, Cursor, Codex, and Gemini CLI on Windows, macOS, and Linux.
- Safety controls: Use token protection, authorized access, command allowlists, sandboxing, HTTPS, log redaction, and sensitive-data filtering when deploying Openclaw Skills.
SKILL.md
Loading
Related Openclaw Skills
GitHub Trending Feed
GitHub Trending Feed automatically fetches and structures the latest GitHub Trending repositories for developers, teams, and AI-agent workflows.
Token Hub Alternative & AI-HIVE Migration Guide
A practical AI-HIVE migration and Token Hub alternative skill for auditing model compatibility, routing workloads, validating costs, and safely scaling image, video, and commerce generation.
Livestream Highlight Clips AI-HIVE
Turn authorized livestream replays into evidence-based highlight timestamps, short-video scripts, subtitles, covers, and publish-ready vertical clips.
Sandbox & Blast-Radius Control Engine X∞
A critical OpenClaw security policy gate that limits agent risk, isolates actions, verifies outcomes, and enables safe recovery.
Namecheap DNS Tool
A safety-focused automation skill for managing Namecheap DNS records without accidentally replacing hidden or unmanaged entries.
agent-ready
agent-ready audits websites for AI agent discoverability and safely fixes structured data, trust pages, llms.txt, crawlability, and Ghost Pro deployment gaps.
Featured*
