Openclaw Skills for Telegram formatting helps AI agents write replies that match Telegram's real rendering rules, avoid broken markup, and stay readable.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install telegram-formatting
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-formatting 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 Telegram Formatting?
Openclaw Skills for Telegram formatting is a developer-focused guide for composing Telegram replies that render correctly in OpenClaw. It distinguishes between standard HTML-based formatting and rich messages mode, so agents can choose the right syntax for the current channel configuration and avoid relying on unsupported tags or broken blocks.
Its core value is practical reliability: use minimal formatting for short replies, structured paragraphs and lists for longer responses, and real Telegram-supported blocks only when they are verified for the current OpenClaw version. Openclaw Skills emphasizes testing, version awareness, and message-shape-driven formatting so responses remain clean across Telegram clients and OpenClaw renderer updates.
Telegram Formatting Use Cases
- Formatting replies in the
telegramchannel with the correct markup for the current mode. - Deciding when to use plain HTML tags like
<b>,<code>,<blockquote>, and links. - Writing rich Telegram messages with headings, tables, and blockquotes when
channels.telegram.richMessagesis enabled. - Preventing broken output from unsupported constructs like
<details>, checklists, pull quotes, and slideshow/collage blocks. - Keeping short replies unformatted and avoiding unnecessary emphasis in one-line confirmations.
- Structuring long troubleshooting answers, status updates, or step-by-step instructions for readability.
- Verifying behavior against the current OpenClaw version before trusting previous formatting test results.
How Telegram Formatting Works
- Check the Telegram rich-message setting once per session with
openclaw config get channels.telegram.richMessages. - Confirm the current OpenClaw version with
openclaw --versionbefore trusting any prior rich-message test results. - If rich messages are disabled or unverified, use Standard mode with HTML parse mode and only supported tags.
- If rich messages are enabled and your version matches the tested version, use plain Markdown-style rich blocks such as headings, tables, and blockquotes, but only in image-free messages.
- Choose formatting based on message shape: no formatting for short replies, light inline tags for medium replies, and paragraphs plus lists for long structured replies.
- Avoid unsupported or low-value constructs, and send separate messages when an image would downgrade the whole reply to a caption.
- Re-test and update the skill file when OpenClaw renderer behavior changes so future replies stay accurate.
Telegram Formatting Setup
- Ensure your OpenClaw environment is available and the Telegram channel is configured.
- Check whether rich messages are enabled:
openclaw config get channels.telegram.richMessages
- Verify your OpenClaw version before relying on any recorded rendering behavior:
openclaw --version
- If
channels.telegram.richMessagesistrue, use rich-message Markdown only for supported blocks and test critical formatting with a throwaway message if your version differs from the tested one. - If rich messages are disabled or uncertain, use Standard mode with HTML tags only.
- When adding or changing formatting behavior, re-test in a real Telegram message and update this skill if the renderer behaves differently.
No additional package installation is required; this skill is a usage and rendering rule set for Openclaw Skills rather than a standalone library.
Telegram Formatting Data Schema & Taxonomy
Openclaw Skills for Telegram formatting does not generate a complex runtime schema, but it follows a strict message-organization model.
| Area | Data / Metadata | Purpose |
|---|---|---|
| Session config | channels.telegram.richMessages |
Determines whether to use Standard mode or rich messages mode. |
| Version metadata | openclaw --version |
Validates whether previously tested formatting behavior is still trustworthy. |
| Message modes | Standard mode, Rich messages mode | Selects the correct rendering strategy for the reply. |
| Standard mode tags | <b>, <i>, <code>, <pre>, <blockquote>, <a> |
Supported HTML formatting for Telegram replies. |
| Rich mode blocks | # headings, Markdown tables, > quote |
Proven-reliable rich blocks in image-free messages. |
| Unsupported blocks | <details>, checklists, pull quotes, slideshow/collage |
Marked unusable or unverified; should not be depended on without retesting. |
| Output constraints | Image caption fallback, 1024-char caption limit | Important transport constraint when a message includes images. |
Taxonomy used by the skill:
- Content length: short, medium, long structured.
- Render mode: standard vs rich.
- Safety status: supported, broken, unverified, or missing feature.
- Media state: image-free vs image-attached, which changes available formatting.
Telegram Formatting Advanced Features
- Session-level config probing to avoid assuming rich-message support.
- Version-aware rendering validation tied to OpenClaw releases.
- Evidence-based formatting guidance that distinguishes tested behavior from documented syntax.
- Image-aware fallback handling to avoid accidental caption downgrades.
- Message-shape heuristics for choosing between no formatting, inline emphasis, or structured blocks.
- Explicit unsupported-feature warnings for collapsible blocks, checklists, pull quotes, and media groups.
- Built-in re-test/update workflow so Openclaw Skills remain accurate as the renderer evolves.
SKILL.md
Loading
Related Openclaw Skills
Editor in Chief
Editor in Chief is an Openclaw Skills editorial agent for correcting prose while preserving the writer’s voice, tone, and intent.
Tailor Resume
Openclaw Skills Tailor Resume updates an existing Google Docs resume for a specific job while staying strictly truthful and evidence-based.
Cloudflare Drop
Cloudflare Drop deploys small static sites to temporary Cloudflare Workers previews with private claim delivery and strict consent controls.
News Impact Scale
News Impact Scale turns any article URL into a personalized prev/current/future impact analysis grounded in historical benchmarks and your own context.
UnlimitedTTS
UnlimitedTTS is an Openclaw Skills workflow for generating paid text-to-speech audio and MP3 narration via MCP tools with x402 USDC settlement.
Volcengine TOS Storage Domain
Openclaw Skills for binding, validating, and troubleshooting public custom domains for Volcengine TOS buckets.