Version 1.2.0 of the WaitingForMacGuffin OpenClaw skill adds platform-aware formatting. When the skill detects it is running inside a Telegram bot, every response type adapts to the constraints of a mobile chat window.
The problem
The skill's default output uses markdown tables, box-drawing tree characters, and wide comparison grids. These render well in Claude desktop and web interfaces but break on Telegram, where markdown tables are not supported and screen width is roughly 40 characters.
A precursor simulation that looks clean on desktop becomes a wall of misaligned text on a phone screen. Data gets lost instead of communicated.
What changed
The skill now includes a full set of Telegram-specific templates for every response type:
- Oscar Brief -- stacked frontrunner list with emoji severity markers instead of a table
- Nominee Deep-Dive -- one data point per line with emoji prefixes instead of tree structures
- Category Overview -- numbered list capped at 5-6 nominees instead of a ranked table
- Bet Picks -- compact card per pick with a quick-compare list instead of a comparison table
- Portfolio -- stacked allocations with a single total line
- Precursor Simulation -- emoji-prefixed lines with a stacked summary
- Risk Legend -- two-line inline legend with dot separators instead of a four-row table
The welcome message is also shorter: seven lines that fit one Telegram screen without scrolling.
General formatting rules for Telegram
- Mobile-first: key numbers front-loaded, ~40 character width assumed
- No markdown tables or box-drawing characters
- One fact per line
- Blank lines as section separators (not horizontal rules)
- Links appended at the end, never inline
Detection
The skill checks for Telegram context automatically: system prompt mentions of Telegram, bot framework identification, or the user stating they are on Telegram. When uncertain, it asks.
Desktop and web users see no change. The default formatting from v1.1 remains the fallback.
How to use it
No action required for existing installations. The skill reads from live APIs and the formatting logic is embedded in the skill instructions, so the update applies immediately.
For new installations:
clawhub install waitingformacguffin
Or grab the skill file from GitHub.