CtrlK
BlogDocsLog inGet started
Tessl Logo
PluginPublic

jbaruch/speaker-toolkit

Twelve skills for conference speakers covering rhetoric analysis, talk authoring, deck building and export, CFP submissions, publishing, illustrations, shownotes, screencast production, and vertical shorts, informed by a 113-entry Presentation Patterns taxonomy (83 observable: 64 patterns + 19 antipatterns; 30 unobservable: 21 patterns + 9 antipatterns).

Quality

92%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Medium

Suggest reviewing before use

Overview
Quality
Evals
Security
Files

humor-post-mortem.mdskills/vault-clarification/references/

Humor Post-Mortem

The skill can identify jokes from transcripts and slides but CANNOT hear laughter. For every talk processed in this run, compile the humor beats detected in dimension 3 and walk through them with the speaker:

  1. List every joke/humor beat identified in the analysis (verbatim quote or slide reference). For each one, ask: "Did this land? Big laugh, knowing nods, or flat?"
  2. Meme slides: For meme-only or meme-with-text slides, ask if the audience visibly reacted or if the meme was more of a visual punchline the speaker talked over.
  3. Spontaneous humor: Ask if there were jokes NOT on the slides that happened in the moment — audience riffs, improvised callbacks, heckler interactions, recovery humor from demo failures. These are invisible to the skill but often the best material.
  4. Humor grading: For each confirmed-landed joke, tag it in the DB with humor_grade: "hit"|"nod"|"flat"|"spontaneous_hit". Over time this builds a corpus-wide humor effectiveness map — which joke TYPES land (self-deprecating, industry snark, meme-as-punchline, callback) and which fall flat.
  5. Promote to portfolio: If a spontaneous joke landed well, ask the speaker if it should be promoted to a planned beat in future deliveries (like the therapy analogy from QCon London 2026).

This is particularly important for recent talks where memory is fresh. For older talks (2+ years), compress to: "Any jokes you remember landing particularly well or badly?"

Store results in humor_postmortem on the talk's DB entry; per-delivery grades stay there and are never appended to the rhetoric summary. Change the summary's Humor & Wit section only when the grades change humor guidance, merging into the existing statement. Persist the DB value only through the ingress owner's update_talk_clarification mutation: name the exact talk filename, set humor_postmortem to one complete JSON object or array, and expect that field's exact current object/array or {"$missing": true}. Dry-run, review, hash-bind apply, and re-read through the owner contract; never edit the tracking database directly.

skills

README.md

tile.json