deck-reviewer

Reviews generated slide images against research sources and storyboard before PowerPoint assembly.

2|1|Updated Mar 7, 2026
One-click install
npx skills add https://github.com/AutoByteus/autobyteus-agents --skill deck-reviewer-autobyteus
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: deck-reviewer
Source: https://github.com/AutoByteus/autobyteus-agents/tree/main/agent-teams/research-to-deck-team/agents/deck-reviewer/skills/deck-reviewer
Command: npx skills add https://github.com/AutoByteus/autobyteus-agents --skill deck-reviewer-autobyteus

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Slide decks built from research often ship with unreadable text, unsupported claims, wrong aspect ratios, or inconsistent styling because no independent QA step exists between slide generation and final assembly. This Skill provides that independent review gate. ## Core Features & Use Cases - Per-slide image inspection: Verifies each slide image is a generate_image or edit_image output, checks 16:9 format, text fidelity, readability, and grounding against the article, claim evidence ledger, and user-approved storyboard. - Structured review report: Produces slide-review-report.md from a template recording pass/fix/reject decisions, exact issues, and required corrections per slide. - Routing and fix loop: Routes slide defects as Deck Fix to infographic_powerpoint_designer and evidence problems as Research Gap to deep_researcher, repeating until no blocking issues remain. - Use Case: After a research-to-deck pipeline generates candidate slide images, run this review to catch a slide with a misquoted statistic and inconsistent typography before final PowerPoint assembly. ## Quick Start Review the candidate slide images in the slides folder against the deck storyboard and claim evidence ledger, then write the slide review report.

Frequently Asked Questions about deck-reviewer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review AI-generated slide images before building a PowerPoint deck?▼

Inspect each slide image against the source article, claim evidence ledger, and approved storyboard, checking text fidelity, aspect ratio, readability, and claim support. Record pass, fix, or reject decisions per slide in a structured review report before assembly.

What should a slide deck QA checklist include?▼

A slide QA checklist should cover 16:9 format, must-appear text presence, no random text or watermarks, claim alignment with evidence, prompt-grounding resource mapping, style and typography consistency, readability at presentation distance, and no sensitive data exposure.

How are slide review issues routed in a research-to-deck workflow?▼

Slide wording, layout, style, readability, or image defects route as Deck Fix to the infographic_powerpoint_designer. Unsupported claims, weak evidence, or missing source anchors route as Research Gap to the deep_researcher, and the loop repeats until no blocking issues remain.

When should slide review reject a generated image?▼

Reject or block slides with missing self-check evidence, unreadable or wrong text, unsupported claims, unindexed source images, script-composed edits instead of generate_image or edit_image outputs, wrong aspect ratio, or random generated text.

What inputs does slide image review require?▼

Review requires the article, research resource index, claim evidence ledger, user-approved deck storyboard, slides visual plan, prompts, deck source index, slide generation log, and the candidate slide images. Missing inputs route back to the deck designer.