CtrlK
BlogDocsLog inGet started
Tessl Logo

jbaruch/speaker-toolkit

Four-skill presentation system: ingest talks into a rhetoric vault, run interactive clarification, generate a speaker profile, then create new presentations that match your documented patterns. Includes an 88-entry Presentation Patterns taxonomy for scoring, brainstorming, and go-live preparation.

96

1.21x
Quality

93%

Does it follow best practices?

Impact

97%

1.21x

Average score across 30 eval scenarios

SecuritybySnyk

Advisory

Suggest reviewing before use

Overview
Quality
Evals
Security
Files

task.mdevals/scenario-1/

Guardrail Audit for a Draft Conference Talk

Background

Morgan Lee is a frequent conference speaker who has just finished a first draft of a 45-minute talk for DevRelCon 2025. Morgan wants a structured quality audit of the draft before presenting it to their co-author. The audit should check every standard guardrail category and flag any violations.

Output Specification

Produce a guardrail report saved to guardrail-report.md covering all standard check categories: slide budget, structure balance, branding, profanity, data attribution, time-sensitive content, closing completeness, cut lines, and anti-patterns. Each check should be labeled with a clear pass/warn/fail verdict. Include a pattern score projection (an estimated numeric score).

Use the speaker profile and draft outline provided below.

Input Files

The following files are provided as inputs. Extract them before beginning.

=============== FILE: inputs/speaker-profile.json =============== { "schema_version": 1, "generated_date": "2025-02-15", "speaker": { "name": "Morgan Lee", "handle": "@mlee_devrel", "shownotes_url_pattern": "speaking.morganlee.io/{slug}" }, "rhetoric_defaults": { "default_duration_minutes": 45, "profanity_calibration": "none", "on_slide_profanity": false, "three_part_close": true, "modular_design": true }, "design_rules": { "footer": { "pattern": "@mlee_devrel | #{conference_hashtag} | speaking.morganlee.io/{slug}" }, "slide_numbers": "never" }, "guardrail_sources": { "slide_budgets": [ {"duration_minutes": 20, "max_slides": 30, "slides_per_min": 1.5}, {"duration_minutes": 30, "max_slides": 45, "slides_per_min": 1.5}, {"duration_minutes": 45, "max_slides": 68, "slides_per_min": 1.5}, {"duration_minutes": 60, "max_slides": 90, "slides_per_min": 1.5} ], "act1_ratio_limits": [ {"duration_minutes": 45, "max_percentage": 45} ], "recurring_issues": [ { "issue": "meme_accretion", "description": "Act 1 accumulates too many meme/image-only slides before the core argument begins", "guardrail": "If Act 1 has more than 60% meme/image-only slides, flag it", "severity": "warn" }, { "issue": "theoretical_framing_delay", "description": "Opens with a long theoretical context section before getting to the practical point", "guardrail": "If the opening framework section exceeds 10% of total slides before any concrete example, flag it", "severity": "warn" } ] } } =============== END OF FILE ===============

=============== FILE: inputs/draft-outline.md ===============

Developer Relations in the Age of AI Assistants

Spec: practitioner | 45 min | DevRelCon 2025 | DevRel professionals and community managers Slide budget: 68 slides (TBD)

Opening Sequence [5 min, slides 1-7]

Slide 1: Title Slide

Slide 2: Bio

Slide 3: Opening Meme — "AI is eating the world"

Slide 4: Opening Meme — Developer confusion about AI tools

Slide 5: Opening Meme — "Documentation? We don't do that here"

Slide 6: Opening Meme — The "just ask ChatGPT" meme

Slide 7: Shownotes URL

Act 1: The Challenge [18 min, slides 8-33]

Slide 8: What is DevRel anyway?

  • Body: Three bullet points defining DevRel
  • Speaker: "Before we get into AI, let's align on definitions."

Slide 9: History of DevRel (2000-2010)

Slide 10: History of DevRel (2010-2015)

Slide 11: History of DevRel (2015-2020)

Slide 12: History of DevRel (2020-2024)

Slide 13: The Tooling Landscape

Slide 14: Traditional Documentation Workflows

Slide 15: The Documentation Gap

Slide 16: Survey Result — "84% of developers say docs are outdated"

  • No source listed

Slide 17: Survey Result — "71% have given up on a product due to bad docs"

  • No source listed

Slide 18: "So what do developers do instead?"

Slide 19: The AI Assistant Adoption Curve

Slide 20: Meme — "Stack Overflow vs. ChatGPT"

Slide 21: Data: AI tool adoption stats (2024)

  • Source: "Various reports"

Slide 22: "The DevRel Fear Response"

  • Body: 6 bullet points: • Fear #1... • Fear #2... • Fear #3... • Fear #4... • Fear #5... • Fear #6...

Slide 23: Fear #1 in detail

Slide 24: Fear #2 in detail

Slide 25: Fear #3 in detail

Slide 26: Fear #4 in detail

Slide 27: Fear #5 in detail

Slide 28: Fear #6 in detail

Slide 29: "But are these fears justified?"

Slide 30-33: Supporting data slides (various)

Act 2: The Opportunity [17 min, slides 34-56]

Slide 34: "Here's the reframe"

Slide 35-56: Solutions and case studies

Closing Sequence [3 min, slides 57-60]

Slide 57: Summary — 3 key takeaways

Slide 58: Call to Action

Slide 59: Shownotes + QR

Slide 60: Thanks

Total slides: 60 =============== END OF FILE ===============

evals

scenario-1

criteria.json

task.md

README.md

tile.json