Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Review MDPR Markdown presentation workflows with semantic hints, visual checks, and deterministic renderer boundaries.
.claude/skills/sickn33-mdpr-skill/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-15 | ✗→✓ | ▲ Improved | 92% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 136% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 33% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -21% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 20% | 0% |
Use this skill as the optional agent companion for MDPR, a deterministic Markdown-to-presentation runtime. MDPR owns parsing, layout, theming, validation, and final PPTX/HTML/PDF rendering. This skill helps an agent review MDPR workflows, propose weak semantic hints, and explain visual findings without taking control of slide geometry.
The upstream skill source is ch040602/mdpr-skill, which includes schemas, review commands, compatibility artifacts, visual evidence examples, and MDPR boundary documentation.
mdpresent, Markdown-to-PPTX, orMarkdown presentation review.
visual review notes.
MDPR as the deterministic renderer.
codex-ppt style workflow.
approval-bound MDPR candidate instead of direct final slide edits.
geometry, typography, colors, z-order, arrows, effects, exact icon assets, renderer object IDs, and final PPTX objects.
deterministic policy changes, or approval-bound proposals.
source draft.
Classify the user's request before producing advice:
semantic hints: compact intent, grouping, importance, and icon-keywordsuggestions.
review report: visual or narrative concerns grounded in rendered evidence,manifests, or validation reports.
layout intent: high-level layout goals from a summarized template catalog,never concrete placeholder coordinates.
theme candidate: reusable token and style-pack proposal for later MDPRapproval/import gates.
codex-ppt compatibility: feature mapping and comparison notes only; do notturn MDPR into a full-slide image renderer.
Reference available evidence such as:
agent-hint.json, review-report.json, ormdpr-theme-candidate-v1
If evidence is missing, say what artifact is needed instead of inventing a pass/fail result.
Allowed hints:
semantically ambiguous
Disallowed hints:
When repeated issues appear, recommend a deterministic follow-up surface:
Run these only when the upstream mdpr-skill CLI is available in the current workspace and the referenced input files exist.
bashnode bin/mdpr-skill.js hint --source-sha256 <64hex> --out .mdpresent/proposals/agent-hint.json node bin/mdpr-skill.js review --manifest dist/mdpresent-manifest.json --out .mdpresent/review/review-report.json node bin/mdpr-skill.js narrative --markdown deck.md --manifest dist/mdpresent-manifest.json --out .mdpresent/review/narrative-review.json node bin/mdpr-skill.js layout-intent --layout-catalog template-layout-catalog.json --out .mdpresent/review/layout-intent.json node bin/mdpr-skill.js accessibility --markdown deck.md --audience "executive review" --out .mdpresent/review/accessibility-review.json
validation report.
markdownFinding: Slide 4 has weak visual hierarchy between the metric and explanation. Evidence: rendered/slide-04.png, manifest slide id `s4`, heading "Revenue Mix". MDPR-owned fix: adjust the metric-card recipe spacing rule or choose a deterministic layout profile with stronger numeric emphasis.
best-fit scenarios.
mdpr-theme-candidate-v1.mdprOwnsFinalLayout, mdprOwnsFinalThemeBinding, andnoRawUseInAgentHints true.
json{ "schema": "mdpr-theme-candidate-v1", "source": "rendered reference set approved by user", "useCases": ["executive review", "research update"], "constraints": { "mdprOwnsFinalLayout": true, "mdprOwnsFinalThemeBinding": true, "noRawUseInAgentHints": true } }
Use codex-ppt only as a capability reference or image-only baseline. Preserve the output-model distinction: codex-ppt style workflows may produce full-slide images, while MDPR defaults to editable PPTX/HTML/PDF with deterministic validation.
markdownComparison note: codex-ppt style output may optimize for a single rasterized slide image. MDPR should instead preserve editable slide objects and route visual improvements through recipes, themes, and validation policies.
gate.
agent judgment alone.
slide content.
Solution: Keep hints semantic and let MDPR choose final layout, geometry, and renderer objects.
Solution: Link each finding to source Markdown, a manifest entry, rendered previews, validation reports, or another concrete artifact.
Solution: Use image-only generators as comparison baselines while preserving MDPR's editable PPTX/HTML/PDF output model.
permission.
review reports or theme candidates.
before running them.
@frontend-slides - Use for browser-native HTML presentation generation.@2slides-ppt-generator - Use for hosted API-based presentation generation.@office-productivity - Use for broader document, spreadsheet, and slideworkflow coordination.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-15 | fail→pass | 10,498 | 6,813 | -35% | 1 | 1 | 0% | 1,700 | 3,270 | +92% | 0 | 0 | — |
case-11 | pass→pass | 15,911 | 11,806 | -26% | 1 | 1 | 0% | 2,599 | 3,854 | +48% | 0 | 0 | — |
case-12 | pass→pass | 7,417 | 6,095 | -18% | 1 | 1 | 0% | 1,229 | 3,007 | +145% | 0 | 0 | — |
case-13 | pass→pass | 7,763 | 6,490 | -16% | 1 | 1 | 0% | 1,253 | 2,895 | +131% | 0 | 0 | — |
case-14 | fail→pass | 7,680 | 6,818 | -11% | 1 | 1 | 0% | 1,291 | 3,052 | +136% | 0 | 0 | — |
case-04 | fail→pass | 27,954 | 26,434 | -5% | 1 | 1 | 0% | 6,186 | 8,202 | +33% | 0 | 0 | — |
case-05 | pass→pass | 12,738 | 10,367 | -19% | 1 | 1 | 0% | 2,500 | 4,055 | +62% | 0 | 0 | — |
case-01 | pass→pass | 22,705 | 13,872 | -39% | 1 | 1 | 0% | 3,898 | 4,354 | +12% | 0 | 0 | — |
case-02 | fail→pass | 28,375 | 11,799 | -58% | 1 | 1 | 0% | 5,307 | 4,183 | -21% | 0 | 0 | — |
case-03 | pass→pass | 18,101 | 8,585 | -53% | 1 | 1 | 0% | 3,231 | 3,844 | +19% | 0 | 0 | — |
case-06 | fail→pass | 10,684 | 3,949 | -63% | 1 | 1 | 0% | 2,222 | 2,663 | +20% | 0 | 0 | — |
case-07 | fail→pass | 5,690 | 2,313 | -59% | 1 | 1 | 0% | 1,048 | 2,448 | +134% | 0 | 0 | — |
case-08 | fail→pass | 7,557 | 2,068 | -73% | 1 | 1 | 0% | 1,384 | 2,368 | +71% | 0 | 0 | — |
case-09 | fail→pass | 5,469 | 2,700 | -51% | 1 | 1 | 0% | 1,102 | 2,495 | +126% | 0 | 0 | — |
case-10 | pass→pass | 9,350 | 5,373 | -43% | 1 | 1 | 0% | 1,505 | 3,000 | +99% | 0 | 0 | — |
case-16 | pass→pass | 8,187 | 2,869 | -65% | 1 | 1 | 0% | 1,334 | 2,587 | +94% | 0 | 0 | — |
case-17 | pass→pass | 10,730 | 4,067 | -62% | 1 | 1 | 0% | 1,751 | 2,684 | +53% | 0 | 0 | — |
case-18 | pass→pass | 9,875 | 4,948 | -50% | 1 | 1 | 0% | 1,850 | 2,960 | +60% | 0 | 0 | — |
case-19 | fail→pass | 6,335 | 2,924 | -54% | 1 | 1 | 0% | 1,213 | 2,589 | +113% | 0 | 0 | — |
case-20 | pass→pass | 11,406 | 5,771 | -49% | 1 | 1 | 0% | 1,930 | 3,079 | +60% | 0 | 0 | — |
case-21 | fail→pass | 12,815 | 8,528 | -33% | 1 | 1 | 0% | 2,153 | 3,508 | +63% | 0 | 0 | — |
case-22 | pass→pass | 10,303 | 4,806 | -53% | 1 | 1 | 0% | 1,908 | 2,973 | +56% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +45 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.