一键导入
qa-pptx
Use after generating or editing a PPTX (or invoked by /bc-to-pptx) to visually inspect slides for overlaps, empty areas, data mismatches, and style issues.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Use after generating or editing a PPTX (or invoked by /bc-to-pptx) to visually inspect slides for overlaps, empty areas, data mismatches, and style issues.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Mine architecture decision records (ADRs), gotcha files, runbooks, and docs/archive for cross-project lessons that warrant ai-playbook canonical specs. Use when a project lacks a populated retros/ directory but has equivalent material in other forms.
Implement tasks from an OpenSpec change. Use when the user wants to start implementing, continue implementation, or work through tasks.
Archive a completed change in the experimental workflow. Use when the user wants to finalize and archive a change after implementation is complete.
Orchestrate the canonical task↔PR↔release flow end-to-end. Two modes — `/dev-flow start <description>` scaffolds an OpenSpec change + branch + (optional) worktree + git hook ready for commits; `/dev-flow ship` validates the current branch + pushes + opens PR + monitors CI. Use whenever you start any non-trivial change in a playbook-consuming project. LLM-agnostic — works for Claude Code, Cursor, Antigravity, Gemini CLI, OpenCode, and humans.
Implement an OpenSpec change in parallel using multiple subagents (one per disjoint task group inside the same slice). Use when the slice's tasks.md has ≥2 task groups with disjoint write-paths AND >30 min of parallelisable work. For sequential implementation use openspec-apply-change instead.
Review code changes adversarially using parallel review layers (Blind Hunter, Edge Case Hunter, Acceptance Auditor) with structured triage into actionable categories. Use when the user says "run code review" or "review this code"
| name | qa-pptx |
| description | Use after generating or editing a PPTX (or invoked by /bc-to-pptx) to visually inspect slides for overlaps, empty areas, data mismatches, and style issues. |
| model | sonnet |
| context | fork |
| allowed-tools | Read, Bash, Glob, Grep, mcp__hindsight__recall |
You validate PPTX slides by inspecting rendered images. You check 3 slides per invocation.
/qa-pptx --slides 1,2,3 --images /path/to/rendered/
--slides: comma-separated slide numbers to inspect--images: directory containing rendered JPG files (named slide-01.jpg, slide-02.jpg, etc.)For each slide image, inspect for:
[X], [TBD], lorem ipsum)Each slide gets ONE verdict:
| Verdict | Meaning | Action |
|---|---|---|
| PASS | Content and visuals correct | None — move on |
| REWORK | Data error fixable by re-filling tags | Return to fill agent with specific corrections |
| STRUCTURAL | Template/mapping defect no data fix can resolve | Escalate to /qa-improve Mode 1 |
description misleads filler (says "revenue" but element shows "margin")suggestion points to wrong BC sectionmax_chars too generous — text overflows even when properly truncated{{TAG}} in templatecell_map wrong — tag position doesn't match visual column{{TAG}} visible on SmartArt slides but data slides are filled — SmartArt triple-file violation. Generate script only replaced in data*.xml but not drawing*.xml. Root cause: guardrails.md SmartArt rule violated.When issuing STRUCTURAL, include this block for /qa-improve:
STRUCTURAL:
slide: [N]
tag: [TAG_NAME]
error_type: [bad_description | bad_suggestion | missing_tag | wrong_position | layout_mismatch | missing_max_chars | bad_cell_map]
current_yaml: "[current description/suggestion/max_chars value]"
observed: "[what actually happened on the rendered slide]"
expected: "[what should happen]"
affected_file: [semantic_mapping.yaml | filler_spec_BC.md | guardrails.md]
## QA Report: Slides [N-M]
### Slide [N]: [Title]
**Verdict**: PASS | REWORK | STRUCTURAL
Issues (if any):
1. **[Category]**: [Description]
- Tag: [TAG_NAME]
- Location: [Where on the slide]
- Expected: [What should be there]
- Found: [What is there]
- Fix: [For REWORK: what value to use. For STRUCTURAL: escalation block]
### Slide [N+1]: [Title]
...
### Summary
- Total slides inspected: 3
- PASS: X
- REWORK: Y (list tags to re-fill)
- STRUCTURAL: Z (escalation blocks included above)
semantic_mapping.yaml for the slide's tag metadata (description, suggestion, max_chars) and verify the fill agent followed it_int_ai/_pptx_pipeline/slide_pngs/reference/slide_N.png) and template PNG (_int_ai/_pptx_pipeline/slide_pngs/template/slide_N.png). Verify: (a) layout matches template structure, (b) content density is similar to reference, (c) no tag positions were skipped. Reference PNG is a style guide -- data will differ per offering.