CtrlK
BlogDocsLog inGet started
Tessl Logo

jbaruch/speaker-toolkit

Two-skill presentation system: analyze your speaking style into a rhetoric knowledge vault, then create new presentations that match your documented patterns. Includes an 88-entry Presentation Patterns taxonomy for scoring, brainstorming, and go-live preparation.

96

1.57x

Quality

96%

Does it follow best practices?

Impact

96%

1.57x

Average score across 15 eval scenarios

Overview
Skills
Evals
Files

rubric.jsonevals/scenario-8/

{
  "context": "Tests whether the agent follows the CFP abstract writing workflow from the skill: lighter Phase 0-1, skips Phase 2, and produces the required deliverables (title, abstract within word count, takeaways count, and speaker bio).",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Abstract word count",
      "description": "The abstract is between 200 and 300 words (inclusive)",
      "max_score": 10
    },
    {
      "name": "Takeaway count",
      "description": "The submission includes between 3 and 5 key takeaways as bullet points",
      "max_score": 10
    },
    {
      "name": "Speaker bio included",
      "description": "A speaker bio is included, drawing from the profile's bio_short and bio_context fields",
      "max_score": 8
    },
    {
      "name": "Title present",
      "description": "The submission includes a talk title that relates to the AI code assistants topic",
      "max_score": 6
    },
    {
      "name": "Presentation spec produced",
      "description": "A separate presentation spec document is produced with at least: title, thesis, audience, venue, and mode fields",
      "max_score": 10
    },
    {
      "name": "Spec includes mode",
      "description": "The presentation spec references a mode from the speaker profile (e.g., 'Myth Buster') rather than inventing a new mode",
      "max_score": 8
    },
    {
      "name": "No full architecture",
      "description": "The output does NOT include a full rhetorical architecture (no instrument selection menu, no slide budget, no outline sections) — the CFP workflow skips the architecture phase",
      "max_score": 10
    },
    {
      "name": "Speaker voice in abstract",
      "description": "The abstract reflects the speaker's voice — contains at least one of: a provocative framing, a bold claim, or a myth-busting angle (not generic academic prose)",
      "max_score": 8
    },
    {
      "name": "Spec includes duration",
      "description": "The presentation spec captures the duration target (30 or 45 minutes based on the conference slot options)",
      "max_score": 6
    },
    {
      "name": "Spec includes profanity register",
      "description": "The presentation spec includes a profanity register setting (informed by the per_audience calibration from the profile and the DevRel audience)",
      "max_score": 6
    },
    {
      "name": "Spec includes commercial intent",
      "description": "The presentation spec includes a commercial intent field",
      "max_score": 6
    },
    {
      "name": "Spec includes shownotes slug",
      "description": "The presentation spec includes a shownotes slug derived from the speaker profile's URL pattern (e.g., sam.codes/ai-code-assistants or similar)",
      "max_score": 6
    },
    {
      "name": "Anti-sell awareness",
      "description": "The abstract or spec acknowledges the sensitivity of the topic for a DevRel audience (people who promote AI tools), showing awareness of the anti-sell dynamic",
      "max_score": 6
    }
  ]
}

Install with Tessl CLI

npx tessl i jbaruch/speaker-toolkit@0.6.2

evals

README.md

tile.json