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 an illustrated presentation outline with the correct structural elements: Illustration Style Anchor in the header with model and per-format anchors, per-slide Format/Illustration/Image prompt fields, [STYLE ANCHOR] token usage, EXCEPTION justifications, and visual continuity devices — all while maintaining standard outline conventions (spec, slide budget, opening/closing structure).",
"type": "weighted_checklist",
"checklist": [
{
"name": "Illustration Style Anchor section present",
"description": "The outline header contains an 'Illustration Style Anchor' section (## heading) that includes: the style name, a **Model:** line with a backtick-wrapped model name, and at least one '### STYLE ANCHOR' sub-heading with a blockquoted anchor paragraph",
"max_score": 12
},
{
"name": "Multiple format anchors",
"description": "The Style Anchor section contains at least two format-specific anchors (e.g., one for FULL and one for IMG+TXT), each with its own dimensions and anchor paragraph",
"max_score": 8
},
{
"name": "Visual continuity conventions documented",
"description": "The outline header documents visual continuity conventions — either in a dedicated 'Conventions' sub-section or inline within the Style Anchor section. Must mention at least two of: the FIG sequential numbering scheme, the recurring personnel evaluation form, and the deadpan military callout label style",
"max_score": 8
},
{
"name": "Format field on every slide",
"description": "Every slide entry (### Slide N:) includes a '- Format: **FULL**' or '- Format: **IMG+TXT**' or '- Format: **EXCEPTION**' line — no slides are missing the Format field",
"max_score": 12
},
{
"name": "EXCEPTION slides have justification",
"description": "Every slide marked as EXCEPTION includes a justification after the format tag explaining why it uses a real asset (e.g., 'bio slide', 'real screenshot', 'data table')",
"max_score": 8
},
{
"name": "Image prompt uses [STYLE ANCHOR] token",
"description": "At least 80% of non-EXCEPTION slides include an '- Image prompt:' field that starts with `[STYLE ANCHOR]` followed by the specific prompt text — the token references the header anchors",
"max_score": 12
},
{
"name": "Illustration field present",
"description": "Non-EXCEPTION slides include an '- Illustration:' field with a human-readable description of the visual concept, distinct from (and shorter than) the Image prompt",
"max_score": 8
},
{
"name": "Prompt richer than description",
"description": "The Image prompt text is visibly more detailed than the Illustration description for at least 3 slides — including composition details, labeling specifics, or visual elements not in the description",
"max_score": 8
},
{
"name": "Visual continuity: FIG numbering",
"description": "Image prompts reference sequential figure numbering (FIG. 1, FIG. 2, etc.) in at least 5 slides, maintaining the numbering convention from the illustration decisions",
"max_score": 8
},
{
"name": "Visual continuity: progressive form",
"description": "The personnel evaluation form appears as a visual element in at least 2 slides (once partially filled, once more complete) — referenced in the Illustration or Image prompt fields",
"max_score": 8
},
{
"name": "Standard outline structure preserved",
"description": "The outline still includes: spec/budget header, opening sequence with title/bio/shownotes, act structure with time and slide range annotations, [CUT LINE] marker, and three-part closing sequence",
"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