research-gap-analyzer

What it does

The skill performs a runtime scan of project feature documents (docs/features/) and research documents (docs/research/), then identifies research gaps that block achieving high revenue, high LTV, and low churn. It maps KPI targets to required research and prioritizes gaps, automatically initiating deep research for P0/P1 gaps. It registers work in assets/research-status.json, runs a deep-research workflow via providers (openai or google) with a structured markdown request, and saves results with automatically generated frontmatter metadata. It then updates the status and can trigger manifest regeneration via a Bash script.

How it works

1. Step 1: Runtime Scan
   • Scan docs/features//PRD-.md for features.
   • Scan docs/research/*.md for existing research.
   • Do not rely on hard-coded feature lists; always scan latest.
2. Step 2: KPI-Feature-Research Mapping
   • Refer to references/kpi-targets.md to derive necessary research per KPI, mapping to required research items.
3. Step 3: Gap Identification
   • Evaluate research quality with a color/icon scheme (none, concept, complete, update needed).
   • Apply a quality checklist (Executive Summary, quantitative data, Action Items, Japan-First alignment).
4. Step 4: Prioritization
   • Use ROI-based scoring: Priority = (KPI impact × 3) + (ease × 2) + (Japan-First fit × 1).
   • Classify as P0, P1, or P2.
5. Step 5: Status registration + Deep Research
   • Register work in assets/research-status.json before starting.
   • Run deep research with providers (openai or google) following a structured markdown request template.
6. Step 6: Results storage + Auto metadata
   • Save results with automatically generated Frontmatter fields (research_id, title, category, kpi_relevance, feature_domains, competitors, target_apps, target_user_levels, pain_points, learning_topics, actionability, scanner, priority, quality, etc.).
   • Auto-create Frontmatter and trigger manifest regeneration via Bash script (scripts/generate_research_manifest.py).

When to use it

Use when you want to identify critical research gaps blocking KPI achievement and automatically initiate targeted deep research for P0/P1 gaps.

What it can touch

• Tools: claude-code (as declared)
• Files touched: assets/research-status.json, docs/research/*.md (for new results), generated Frontmatter in new Markdown files, references/research-status-schema.json (for schema), and scripts/generate_research_manifest.py (for manifest regeneration).

Caveats

• Frontmatter is auto-generated after deep research; human editing is not required for this step.
• The workflow relies on predefined KPI targets and gap definitions in repository references; outcomes depend on the quality and relevance of input documents.
• The extraction and classification logic is described conceptually; actual implementation details are defined within the skill's automation steps.