UNPKG

aiwg

Version:

Cognitive architecture for AI-augmented software development with structured memory, ensemble validation, and closed-loop correction. FAIR-aligned artifacts, 84% cost reduction via human-in-the-loop, standards adopted by 100+ organizations.

212 lines (158 loc) 7.24 kB
# Research-Complete Framework Templates This directory contains 8 artifact templates for the research-complete framework, following progressive disclosure patterns and AIWG best practices. ## Templates Overview | Template | Format | Use Case | Agent | |----------|--------|----------|-------| | **literature-note.md** | Markdown | Zettelkasten-style note capturing paper insights | Discovery Agent | | **summary.md** | Markdown | Multi-level summary (1-sentence, 1-paragraph, 1-page) | Discovery Agent | | **extraction.yaml** | YAML | Structured extraction of claims, methods, findings | Discovery Agent | | **gap-report.md** | Markdown | Research gap analysis and prioritization | Discovery Agent | | **citation-network.md** | Markdown | Citation network visualization and analysis | Discovery Agent | | **quality-assessment.md** | Markdown | GRADE quality assessment | Quality Agent | | **provenance-record.yaml** | YAML | W3C PROV-compliant provenance tracking | Provenance Agent | | **acquisition-report.md** | Markdown | Paper acquisition with FAIR compliance scores | Acquisition Agent | ## Template Categories ### Note-Taking & Synthesis - `literature-note.md` - Deep reading notes with connections - `summary.md` - Tiered summaries for different audiences ### Analysis & Evaluation - `extraction.yaml` - Machine-readable structured data - `gap-report.md` - Identify unstudied research areas - `citation-network.md` - Map intellectual lineage - `quality-assessment.md` - Evaluate evidence quality with GRADE ### Metadata & Compliance - `provenance-record.yaml` - Track artifact derivation chains - `acquisition-report.md` - Document paper acquisition and verification ## Usage Workflow The recommended workflow for processing a new research paper: ``` 1. acquisition-report.md → Acquire paper, verify identity, check license 2. provenance-record.yaml → Create W3C PROV record 3. literature-note.md → Deep reading, capture insights 4. summary.md → Generate tiered summaries 5. extraction.yaml → Extract structured data 6. quality-assessment.md → Assess evidence quality (GRADE) 7. citation-network.md → Analyze relationships (when corpus sufficient) 8. gap-report.md → Identify gaps (periodic synthesis) ``` ## Key Features ### Progressive Disclosure All markdown templates follow progressive disclosure pattern: - **Phase 1: Core (ESSENTIAL)** - Always visible, complete immediately - **Phase 2: Details (EXPAND WHEN READY)** - Expand during elaboration - **Phase 3: Advanced (ADVANCED)** - Technical details during construction ### Reasoning Sections All templates include mandatory reasoning sections per `@.claude/rules/reasoning-sections.md`: 1. **Problem Analysis** - What are we solving? 2. **Approach Planning** - How will we approach this? 3. **Trade-off Evaluation** - What are the options? 4. **Decision Rationale** - Why this approach? 5. **Impact Assessment** - What are the consequences? ### Worked Examples Every template includes extensive inline examples showing: - **Simple scenarios** - Baseline understanding - **Moderate scenarios** - Realistic complexity - **Complex scenarios** - Edge cases and advanced usage - **Anti-patterns** - What NOT to do (with explanations) ### Schema Compliance YAML templates conform to schemas: - `extraction.yaml``@agentic/code/frameworks/research-complete/schemas/extraction-schema.yaml` - `provenance-record.yaml``@agentic/code/frameworks/sdlc-complete/schemas/provenance/prov-record.yaml` ## Template Selection Guide ### For Literature Review Start with: `literature-note.md` - Deep reading and insight capture - Connect to AIWG implementation needs - Track connections to other papers ### For Quick Reference Start with: `summary.md` - 1-sentence summary for executives - 1-paragraph for practitioners - 1-page for researchers ### For Machine Processing Start with: `extraction.yaml` - Structured claims with evidence - Methods and algorithms - Findings with applicability notes ### For Research Planning Start with: `gap-report.md` - Identify unstudied areas - Prioritize research needs - Plan pilot studies or external research ### For Intellectual Lineage Start with: `citation-network.md` - Map paper relationships - Identify influential work - Find citation paths to AIWG implementations ### For Evidence Quality Start with: `quality-assessment.md` - GRADE quality rating (HIGH/MODERATE/LOW/VERY LOW) - Systematic evaluation - Applicability to AIWG assessment ### For Paper Acquisition Start with: `acquisition-report.md` - Document source and verification - Compute fixity checksums - Assess FAIR compliance ### For Provenance Tracking Start with: `provenance-record.yaml` - W3C PROV-compliant records - Track derivation chains - Attribution and audit trails ## Integration with Research Services These templates are used by research-complete framework services: - **Discovery Service** - Uses literature-note, summary, extraction templates - **Acquisition Service** - Uses acquisition-report, provenance-record templates - **Archival Service** - Uses provenance-record, quality-assessment templates - **Citation Service** - Uses citation-network, extraction templates - **Documentation Service** - Uses summary, gap-report templates - **Provenance Service** - Uses provenance-record template - **Quality Service** - Uses quality-assessment template ## Example Usage ### Create Literature Note ```bash # Using template directly cp templates/literature-note.md .aiwg/research/findings/REF-018-react.md # Or via agent aiwg research note REF-018 --template literature-note ``` ### Generate Multi-Level Summary ```bash # Using template cp templates/summary.md .aiwg/research/summaries/REF-018-summary.md # Or via agent aiwg research summary REF-018 ``` ### Extract Structured Data ```bash # Using template cp templates/extraction.yaml .aiwg/research/extractions/REF-018-extraction.yaml # Or via agent aiwg research extract REF-018 ``` ## Quality Checklist Before completing any artifact from these templates: - [ ] Reasoning section completed FIRST (before content) - [ ] All ESSENTIAL fields populated - [ ] Inline examples provided (not placeholders) - [ ] Anti-patterns documented where applicable - [ ] References include @-mentions to related artifacts - [ ] Progressive disclosure sections properly tagged - [ ] Schema validation passes (for YAML templates) - [ ] Provenance record created (if generating new artifact) ## References - **Framework README**: `@agentic/code/frameworks/research-complete/README.md` - **Agents**: `@agentic/code/frameworks/research-complete/agents/` - **Schemas**: `@agentic/code/frameworks/sdlc-complete/schemas/research/` ### Rules - `@.claude/rules/progressive-disclosure.md` - Template structure - `@.claude/rules/reasoning-sections.md` - Reasoning requirements - `@.claude/rules/provenance-tracking.md` - Provenance tracking - `@.claude/rules/research-metadata.md` - Metadata requirements - `@.claude/rules/citation-policy.md` - Citation standards ## Manifest Full template catalog available in `manifest.json`. --- **Version:** 1.0.0 **Last Updated:** 2026-02-03 **Framework:** research-complete