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
93%
Does it follow best practices?
Impact
97%
1.21xAverage score across 30 eval scenarios
Advisory
Suggest reviewing before use
{
"context": "Tests whether the agent produces a correct PowerPoint deck from an outline and profile: template cleaned of demo slides, correct layout per slide type, footer on every slide, speaker notes present, background color rules respected, and no slide numbers.",
"type": "weighted_checklist",
"checklist": [
{
"name": "Template demo slides removed",
"description": "The output deck does not contain the original demo/sample slides from the template — only the new slides from the outline are present",
"max_score": 12
},
{
"name": "Speaker notes present",
"description": "Slides that have 'Speaker:' lines in the outline have those notes present in the final .pptx file's speaker notes",
"max_score": 20
},
{
"name": "Footer on every slide",
"description": "Every slide in the output deck has a footer text element with the speaker's handle and conference info",
"max_score": 10
},
{
"name": "Layout selection from profile",
"description": "The deck plan maps each slide to a layout from the profile's template_layouts array based on the slide's content type — not hardcoded or random layout choices",
"max_score": 10
},
{
"name": "No adjacent color repeat",
"description": "Background colors on consecutive slides are different — the plan or deck shows no two adjacent slides share the same background color",
"max_score": 10
},
{
"name": "White/black reserved for images",
"description": "White or black backgrounds are only used on full-bleed image or meme slides — content slides get colored backgrounds from the pool",
"max_score": 8
},
{
"name": "No slide numbers",
"description": "The output deck does not include slide numbers (respecting the 'never' rule from the profile)",
"max_score": 6
},
{
"name": "Footer color adapts",
"description": "Footer text color is light on dark backgrounds and dark on light backgrounds — not a fixed color regardless of background",
"max_score": 8
},
{
"name": "Correct slide count",
"description": "The output deck has the same number of slides as the outline (8 slides from the excerpt), not more or fewer",
"max_score": 8
},
{
"name": "Build plan documents decisions",
"description": "The deck_plan.md maps each outline slide to a specific layout and documents the design decisions (color, footer, notes)",
"max_score": 8
}
]
}evals
scenario-1
scenario-2
scenario-3
scenario-4
scenario-5
scenario-6
scenario-7
scenario-8
scenario-9
scenario-10
scenario-11
scenario-12
scenario-13
scenario-14
scenario-15
scenario-16
scenario-17
scenario-18
scenario-19
scenario-20
scenario-21
scenario-22
scenario-23
scenario-24
scenario-25
scenario-26
scenario-27
scenario-28
scenario-29
scenario-30
rules
skills
presentation-creator
references
patterns
build
deliver
prepare
scripts
vault-clarification
vault-ingress
vault-profile