/wait-what — a user-invoked productivity skill, “Stop. That last message did not land: re-pitch it.” Its whole body is one instruction: re-pitch with a little context, in ASD-STE100 Simplified Technical English, using the ubiquitous language from GLOSSARY.md (following GLOSSARY-MAP.md to the right one in a multi-context repo).
It is a “one-word corrective for model verbosity”, usable mid-conversation and inside any other skill.
The mechanism is the name
The changelog argues that concision skills “fail by growing: a 400-line skill still leaves the model verbose.” So this one is three lines, and the work is done by a single leading word (the concept mp-writing-for-agents defines). Names that describe the output (/tldr, /no-fluff) make the model clip words and lose the reader further. Naming the listener’s state asks for both halves at once: fewer words and the missing context.
Where it sits
Off-stage. It repairs one message and does not prevent the next. The upfront cure for jargon is a shared language built with mp-grill-with-docs / mp-domain-modeling (artifact-domain-model). wait-what is the after-the-fact fix that reuses that same vocabulary. It addresses the README’s failure mode #2, “The Agent Is Way Too Verbose”.
See Also
- mp-domain-modeling · artifact-domain-model — the
GLOSSARY.mdit re-pitches in. - mp-writing-for-agents — the leading-word theory it is a three-line instance of.
- mp-ask-matt — routes it under Standalone.
Version history
- v1.3.0 (2026-10-04): follows
GLOSSARY-MAP.mdto the rightGLOSSARY.mdin a multi-context repo; the domain doc it reads was renamed fromCONTEXT.mdtoGLOSSARY.md. - v1.2.0 (2026-08-05): added, user-invoked, three lines long.