Initial commit

skills/synthesis-and-analogy/SKILL.md

@@ -0,0 +1,225 @@
+---
+name: synthesis-and-analogy
+description: Use when synthesizing information from multiple sources (literature review, stakeholder feedback, research findings, data from different systems), creating or evaluating analogies for explanation or problem-solving (cross-domain transfer, "X is like Y", structural mapping), combining conflicting viewpoints into unified framework, identifying patterns across disparate sources, finding creative solutions by transferring principles from one domain to another, testing whether analogies hold (surface vs deep similarities), or when user mentions "synthesize", "combine sources", "analogy", "like", "similar to", "transfer from", "integrate findings", "what's it analogous to".
+---
+# Synthesis & Analogy
+
+## Table of Contents
+- [Purpose](#purpose)
+- [When to Use](#when-to-use)
+- [What Is It](#what-is-it)
+- [Workflow](#workflow)
+- [Synthesis Techniques](#synthesis-techniques)
+- [Analogy Techniques](#analogy-techniques)
+- [Common Patterns](#common-patterns)
+- [Guardrails](#guardrails)
+- [Quick Reference](#quick-reference)
+
+## Purpose
+
+Synthesize information from multiple sources into coherent insights and use analogical reasoning to transfer knowledge across domains, explain complex concepts, and find creative solutions.
+
+## When to Use
+
+**Information Synthesis:**
+- Literature review (combine 10+ research papers into narrative)
+- Multi-source integration (customer feedback + analytics + competitive data)
+- Conflicting viewpoint reconciliation (synthesize disagreeing experts)
+- Pattern identification across sources (themes from interviews, support tickets, reviews)
+
+**Analogical Reasoning:**
+- Explain complex concepts (use familiar domain to explain unfamiliar)
+- Cross-domain problem-solving (transfer solution from different field)
+- Creative ideation (find novel solutions through structural mapping)
+- Teaching/communication (make abstract concepts concrete)
+
+**Combined Synthesis + Analogy:**
+- Synthesize multiple analogies to build richer understanding
+- Use analogies to reconcile conflicting sources ("both are right from different perspectives")
+- Transfer synthesized insights from one domain to another
+
+## What Is It
+
+**Synthesis**: Combining information from multiple sources into unified, coherent whole that reveals patterns, resolves conflicts, and generates new insights beyond individual sources.
+
+**Analogy**: Structural mapping between domains where relationships in source domain (familiar) illuminate relationships in target domain (unfamiliar). Good analogies preserve deep structure, not just surface features.
+
+**Example - Synthesis**: Synthesizing 15 customer interviews + 5 surveys + support ticket analysis → "Customers struggle with onboarding (87% mention), specifically Step 3 configuration (65% abandon here), because terminology is domain-specific (42% request glossary). Three user types emerge: novices (need hand-holding), intermediates (need examples), experts (need speed)."
+
+**Example - Analogy**: "Microservices architecture is like a city of specialized shops vs monolithic architecture like a department store. City: each shop (service) independent, can renovate without closing whole city, but must coordinate deliveries (APIs). Department store: everything under one roof (codebase), easier coordination, but renovating one section disrupts whole store. Trade-off: flexibility vs simplicity."
+
+## Workflow
+
+Copy this checklist and track your progress:
+
+```
+Synthesis & Analogy Progress:
+- [ ] Step 1: Clarify goal and gather sources/domains
+- [ ] Step 2: Choose approach (synthesis, analogy, or both)
+- [ ] Step 3: Apply synthesis or analogy techniques
+- [ ] Step 4: Test quality and validity
+- [ ] Step 5: Refine and deliver insights
+```
+
+**Step 1: Clarify goal**
+
+For synthesis: What sources? What question are we answering? What conflicts need resolving? For analogy: What's source domain (familiar)? What's target domain (explaining)? What's goal (explain, solve, ideate)? See [Common Patterns](#common-patterns) for typical goals.
+
+**Step 2: Choose approach**
+
+Synthesis only → Use [Synthesis Techniques](#synthesis-techniques). Analogy only → Use [Analogy Techniques](#analogy-techniques). Both → Start with synthesis to find patterns, then use analogy to explain or transfer. For straightforward cases → Use [resources/template.md](resources/template.md). For complex multi-domain synthesis → Study [resources/methodology.md](resources/methodology.md).
+
+**Step 3: Apply techniques**
+
+For synthesis: Identify themes across sources, note agreements/disagreements, resolve conflicts via higher-level framework, extract patterns. For analogy: Map structure from source to target (what corresponds to what?), identify shared relationships (not surface features), test mapping validity. See [Synthesis Techniques](#synthesis-techniques) and [Analogy Techniques](#analogy-techniques).
+
+**Step 4: Test quality**
+
+Self-assess using [resources/evaluators/rubric_synthesis_and_analogy.json](resources/evaluators/rubric_synthesis_and_analogy.json). Synthesis checks: captures all sources? resolves conflicts? identifies patterns? adds insight? Analogy checks: structure preserved? deep not surface? limitations acknowledged? helps understanding? Minimum standard: Score ≥3.5 average.
+
+**Step 5: Refine and deliver**
+
+Create `synthesis-and-analogy.md` with: synthesis summary (themes, agreements, conflicts, patterns, new insights) OR analogy explanation (source domain, target domain, mapping table, what transfers, limitations), supporting evidence from sources, actionable implications.
+
+## Synthesis Techniques
+
+**Thematic Synthesis** (identify recurring themes):
+1. **Extract**: Read each source, note key points and themes
+2. **Code**: Label similar ideas with same theme tag (e.g., "onboarding friction", "pricing confusion")
+3. **Count**: Track frequency (how many sources mention each theme?)
+4. **Rank**: Prioritize by frequency × importance
+5. **Synthesize**: Describe each major theme with supporting evidence from sources
+
+**Conflict Resolution Synthesis** (reconcile disagreements):
+- **Meta-level framework**: Both right from different perspectives (e.g., "Source A prioritizes speed, Source B prioritizes quality - depends on context")
+- **Scope distinction**: Disagree on scope ("Source A: feature X broken for enterprise. Source B: works for SMB. Synthesis: works for SMB, broken for enterprise")
+- **Temporal**: Disagreement over time ("Source A: strategy X failed in 2010. Source B: works in 2024. Context changed: market maturity")
+- **Null hypothesis**: Genuinely conflicting evidence → state uncertainty, propose tests
+
+**Pattern Identification** (find cross-cutting insights):
+- Look for repeated structures (same problem in different guises)
+- Find causal patterns (when X, then Y across multiple sources)
+- Identify outliers (sources that contradict pattern - why?)
+- Extract meta-insights (what does the pattern tell us?)
+
+**Example**: Synthesizing 10 postmortems → Pattern: 80% of incidents involve config change + lack of rollback plan. Outliers: 2 incidents hardware failure. Meta-insight: Need config change review process + automatic rollback capability.
+
+## Analogy Techniques
+
+**Structural Mapping Theory**:
+1. **Identify source domain** (familiar, well-understood)
+2. **Identify target domain** (unfamiliar, explaining)
+3. **Map entities**: What in source corresponds to what in target?
+4. **Map relationships**: Preserve relationships (if A→B in source, then A'→B' in target)
+5. **Test mapping**: Do relationships transfer? Are there unmapped elements?
+6. **Acknowledge limits**: Where does analogy break down?
+
+**Surface vs Deep Analogies**:
+- **Surface (weak)**: Share superficial features (both round, both red) - not illuminating
+- **Deep (strong)**: Share structural relationships (both have hub-spoke topology, both use feedback loops) - insightful
+
+**Example - Surface**: "Brain is like computer (both process information)" - too vague, doesn't help
+**Example - Deep**: "Brain neurons are like computer transistors: neurons fire/don't fire (binary), connect in networks, learning = strengthening connections (weights). BUT neurons are analog/probabilistic, computer precise/deterministic" - preserves structure, acknowledges limits
+
+**Analogy Quality Tests**:
+- **Systematicity**: Do multiple relationships map (not just one)?
+- **Structural preservation**: Do causal relations transfer?
+- **Productivity**: Does analogy generate new predictions/insights?
+- **Scope limits**: Where does analogy break? (Always acknowledge)
+
+## Common Patterns
+
+**Pattern 1: Literature Review Synthesis**
+- Goal: Combine research papers into narrative
+- Technique: Thematic synthesis (extract themes, note agreements/conflicts, identify gaps)
+- Output: "Research shows X (5 studies support), but Y remains controversial (3 for, 2 against due to methodology differences). Gap: no studies on Z population."
+
+**Pattern 2: Multi-Stakeholder Synthesis**
+- Goal: Integrate feedback from design, engineering, product, customers
+- Technique: Conflict resolution synthesis (meta-level framework, scope distinctions)
+- Output: "Design wants A (aesthetics), Engineering wants B (performance), Product wants C (speed). All valid - prioritize C (speed) for v1, A (aesthetics) for v2, B (performance) as ongoing optimization."
+
+**Pattern 3: Explanatory Analogy**
+- Goal: Explain technical concept to non-technical audience
+- Technique: Structural mapping from familiar domain
+- Output: "Git branches are like alternate timelines in sci-fi: main branch is prime timeline, feature branches are 'what if' explorations. Merge = timeline convergence. Conflicts = paradoxes to resolve."
+
+**Pattern 4: Cross-Domain Problem-Solving**
+- Goal: Solve problem by transferring solution from different field
+- Technique: Identify structural similarity, map solution elements
+- Output: "Warehouse routing problem is structurally similar to ant colony optimization: ants find shortest paths via pheromone trails. Transfer: use reinforcement learning with 'digital pheromones' (successful route weights) to optimize warehouse paths."
+
+**Pattern 5: Creative Ideation via Analogy**
+- Goal: Generate novel ideas by exploring analogies
+- Technique: Forced connections, random domain pairing, systematic variation
+- Output: "How is code review like restaurant food critique? Critic (reviewer) evaluates dish (code) on presentation (readability), taste (correctness), technique (architecture). Transfer: multi-criteria rubric for code review focusing on readability, correctness, architecture."
+
+## Guardrails
+
+**Synthesis Quality:**
+- Covers all relevant sources (no cherry-picking)
+- Resolves conflicts explicitly (doesn't ignore disagreements)
+- Identifies patterns beyond what individual sources state (adds value)
+- Distinguishes facts from interpretations
+- Cites sources for claims
+- Acknowledges gaps and uncertainties
+
+**Analogy Quality:**
+- Maps structure not surface features (deep analogy)
+- Explicitly states what corresponds to what (mapping table)
+- Tests validity (do relationships transfer?)
+- Acknowledges where analogy breaks down (limitations)
+- Doesn't overextend (knows when to stop pushing analogy)
+- Appropriate for audience (familiar source domain)
+
+**Avoid:**
+- **False synthesis**: Forcing agreement where genuine conflict exists
+- **Surface analogies**: "Both are round" doesn't help understanding
+- **Analogy as proof**: Analogies illustrate, don't prove
+- **Overgeneralization**: One source ≠ pattern
+- **Cherry-picking**: Ignoring inconvenient sources
+- **Mixing levels**: Confusing data with interpretation
+
+## Quick Reference
+
+**Inputs Required:**
+
+For synthesis:
+- Multiple sources (papers, interviews, datasets, feedback, research)
+- Question to answer or goal to achieve
+- Conflicts or patterns to identify
+
+For analogy:
+- Source domain (familiar, well-understood)
+- Target domain (unfamiliar, explaining or solving)
+- Goal (explain, solve problem, generate ideas)
+
+**Techniques to Use:**
+
+Synthesis:
+- Thematic synthesis → Identify recurring themes
+- Conflict resolution → Reconcile disagreements via meta-framework
+- Pattern identification → Find cross-cutting insights
+
+Analogy:
+- Structural mapping → Map entities and relationships
+- Surface vs deep test → Ensure structural not superficial similarity
+- Validity test → Check if relationships transfer
+
+**Outputs Produced:**
+
+- `synthesis-and-analogy.md` with:
+  - Synthesis: themes, agreements, conflicts resolved, patterns, new insights, supporting evidence
+  - Analogy: source domain, target domain, mapping table (what↔what), transferred insights, limitations
+  - Actionable implications
+
+**Resources:**
+- Quick synthesis or analogy → [resources/template.md](resources/template.md)
+- Complex multi-source or multi-domain → [resources/methodology.md](resources/methodology.md)
+- Quality validation → [resources/evaluators/rubric_synthesis_and_analogy.json](resources/evaluators/rubric_synthesis_and_analogy.json)
+
+**Minimum Quality Standard:**
+- Synthesis: covers all sources, resolves conflicts, identifies patterns, adds insight
+- Analogy: structural mapping clear, deep not surface, limitations acknowledged
+- Both: evidence-based, cited sources, actionable
+- Average rubric score ≥ 3.5/5 before delivering

skills/synthesis-and-analogy/resources/evaluators/rubric_synthesis_and_analogy.json

@@ -0,0 +1,146 @@
+{
+  "criteria": [
+    {
+      "name": "Source Coverage & Completeness (Synthesis)",
+      "weight": 1.3,
+      "description": "For synthesis: Are all relevant sources included and documented?",
+      "levels": {
+        "5": "All relevant sources systematically included and documented. Source inventory complete with type, key claims, quality assessment. No cherry-picking. Actively seeks disconfirming evidence. Source quality explicitly assessed (high/medium/low). N/A if analogy-only.",
+        "4": "Most relevant sources included. Source inventory mostly complete. Minor sources may be missing. Source quality noted for most. Mostly avoids cherry-picking.",
+        "3": "Major sources included but some gaps. Basic source documentation. Some cherry-picking possible (only confirms pre-existing view). Source quality mentioned inconsistently.",
+        "2": "Significant source gaps. Incomplete documentation. Clear cherry-picking (ignores disconfirming sources). Source quality not assessed.",
+        "1": "Few sources or only sources supporting one view. No systematic documentation. Heavy cherry-picking. Source quality ignored."
+      }
+    },
+    {
+      "name": "Thematic Analysis & Pattern Identification (Synthesis)",
+      "weight": 1.4,
+      "description": "For synthesis: Are themes and patterns identified with evidence?",
+      "levels": {
+        "5": "3-5 major themes identified with frequency counts (X/Y sources mention). Patterns supported by 3+ sources with clear evidence. Distinguishes themes (single topic) from patterns (repeated structure). New insights generated beyond individual sources. N/A if analogy-only.",
+        "4": "Major themes identified with some frequency data. Patterns noted with supporting evidence from multiple sources. Some new insights. Minor distinction issues between themes and patterns.",
+        "3": "Themes identified but minimal frequency data. Patterns claimed without sufficient evidence (maybe 1-2 sources). Limited new insights (mostly summarizes existing sources).",
+        "2": "Vague themes without evidence. Single-source 'patterns' (overgeneralization). No new insights (just combines existing claims without synthesis).",
+        "1": "No thematic analysis or pattern identification. Just lists source claims without integration."
+      }
+    },
+    {
+      "name": "Conflict Resolution (Synthesis)",
+      "weight": 1.2,
+      "description": "For synthesis: Are conflicts between sources explicitly addressed and resolved?",
+      "levels": {
+        "5": "Conflicts between sources explicitly identified. Resolution provided using meta-framework (both right from different perspectives), scope distinction (disagree on scope), temporal distinction (changed over time), OR uncertainty stated (genuinely conflicting). Doesn't force false consensus. N/A if analogy-only or no conflicts.",
+        "4": "Conflicts identified and addressed. Resolution mostly clear. May paper over minor conflicts. Generally avoids false consensus.",
+        "3": "Some conflicts noted but resolution unclear or incomplete. May ignore some disagreements. Partial false consensus.",
+        "2": "Conflicts ignored or dismissed without justification. Forces agreement where genuine disagreement exists. False synthesis.",
+        "1": "Doesn't acknowledge conflicts despite clear disagreements between sources. Presents as if all sources agree."
+      }
+    },
+    {
+      "name": "Structural Mapping Quality (Analogy)",
+      "weight": 1.5,
+      "description": "For analogy: Is the structural mapping between domains explicit and valid?",
+      "levels": {
+        "5": "Explicit mapping table showing entity correspondences (A↔X) and relationship correspondences (A→B maps to X→Y). Maps relational structure, not surface attributes. Multiple interconnected relations mapped (systematicity). Mapping validity tested. N/A if synthesis-only.",
+        "4": "Clear mapping between domains. Mostly relational (some attribute mapping). Several relations mapped. Mapping generally valid.",
+        "3": "Basic mapping provided but may mix relational and attribute-based. Limited relations mapped (1-2). Validity assumed not tested.",
+        "2": "Vague mapping ('A is like X'). Primarily surface/attribute-based ('both are round'). Single relation or unclear. Not validated.",
+        "1": "No explicit mapping or purely surface similarity. No relational structure. Unclear what corresponds to what."
+      }
+    },
+    {
+      "name": "Analogy Depth & Systematicity (Analogy)",
+      "weight": 1.4,
+      "description": "For analogy: Is it a deep (relational) analogy with multiple interconnected mappings?",
+      "levels": {
+        "5": "Deep analogy mapping relational structure (A→B in source maps to X→Y in target with same causal/structural relationship). Systematic: 3+ interconnected relations map. Higher-order relations (relations between relations) map. Not just surface features. N/A if synthesis-only.",
+        "4": "Mostly deep analogy with relational mapping. 2-3 relations mapped with some interconnection. Higher-order relations may be implied.",
+        "3": "Mix of deep and surface. Some relational mapping but also attribute-based. 1-2 relations mapped. Limited interconnection.",
+        "2": "Primarily surface analogy based on attributes ('both process information' - vague). Minimal relational structure. Isolated facts, not interconnected system.",
+        "1": "Pure surface analogy ('both are blue', 'both involve computers'). No relational mapping. Doesn't help understanding."
+      }
+    },
+    {
+      "name": "Productivity & Transferability (Analogy)",
+      "weight": 1.3,
+      "description": "For analogy: Does it generate new predictions or transfer useful insights?",
+      "levels": {
+        "5": "Analogy generates new predictions or insights about target domain not previously obvious. Insights transferred from source are specific and actionable. Testable predictions made. Explains mechanism (not just correlation). N/A if synthesis-only.",
+        "4": "Generates some new insights. Transfer is somewhat actionable. Predictions or explanations provided but may be less specific.",
+        "3": "Limited new insights (mostly restates what's known). Transfer is vague ('could help with X'). No specific predictions.",
+        "2": "No new insights (just restates target in source domain terms without adding understanding). Transfer unclear. No predictions.",
+        "1": "Adds no understanding. Analogy obscures rather than clarifies. Nothing transfers usefully."
+      }
+    },
+    {
+      "name": "Scope Limitations Acknowledged",
+      "weight": 1.2,
+      "description": "For analogy: Are limitations explicitly stated (where analogy breaks down)?",
+      "levels": {
+        "5": "Limitations explicitly and specifically stated: 'Analogy holds for X, Y, Z but breaks down for A, B, C because [reasons].' Knows when to stop pushing analogy. Prevents overextension. CRITICAL for all analogies.",
+        "4": "Limitations noted ('analogy has limits') with some specifics. Generally avoids overextension.",
+        "3": "Generic limitation statement ('no analogy is perfect') without specifics. May push analogy slightly beyond valid scope.",
+        "2": "Minimal acknowledgment of limitations. Overextends analogy. Doesn't specify where it breaks.",
+        "1": "No limitations acknowledged. Treats analogy as if it holds perfectly. Dangerous overextension."
+      }
+    },
+    {
+      "name": "Evidence & Citation Quality",
+      "weight": 1.1,
+      "description": "Are claims supported by evidence? Are sources cited?",
+      "levels": {
+        "5": "All major claims backed by evidence (quotes, data, source citations). Facts distinguished from interpretations ('Source A shows X [fact]. This suggests Y [interpretation]'). Sources properly cited (author, title, date or source #). Evidence strength assessed (high/medium/low confidence).",
+        "4": "Most claims backed by evidence. Sources generally cited. Mostly distinguishes facts from interpretations. Some evidence strength noted.",
+        "3": "Some claims backed by evidence, others asserted. Inconsistent citations. Facts and interpretations sometimes blurred.",
+        "2": "Minimal evidence for claims. Few citations. Assertions presented as facts. Evidence strength not assessed.",
+        "1": "No evidence or citations. Pure assertions. Speculation presented as fact."
+      }
+    }
+  ],
+  "guidance": {
+    "task_type": {
+      "synthesis_only": "Prioritize Source Coverage (1.3x), Thematic Analysis (1.4x), Conflict Resolution (1.2x). Structural Mapping, Analogy Depth, Productivity, and Limitations criteria are N/A (not scored). Focus on comprehensive source integration, pattern identification, and conflict reconciliation.",
+      "analogy_only": "Prioritize Structural Mapping (1.5x), Analogy Depth (1.4x), Productivity (1.3x), Limitations (1.2x). Source Coverage, Thematic Analysis, and Conflict Resolution criteria are N/A (not scored). Focus on deep relational mapping, systematicity, transferable insights, and explicit scope limits.",
+      "both_synthesis_and_analogy": "All criteria apply. Common pattern: synthesize first (identify patterns across sources), then use analogy to explain synthesized insights to audience (map pattern to familiar domain). Both synthesis and analogy work should meet quality standards."
+    },
+    "complexity": {
+      "simple": "Simple synthesis (3-5 similar sources, clear consensus) or simple analogy (familiar domains, obvious mapping). Target score ≥3.5 average. All criteria ≥3.",
+      "moderate": "Moderate complexity (5-10 sources with some conflicts) or moderate analogy (less familiar domain, multiple relations to map). Target score ≥4.0 average. All criteria ≥3.",
+      "complex": "Complex synthesis (10+ sources, significant conflicts, cross-domain/temporal) or complex analogy (abstract domains, higher-order relations, multiple interconnected mappings). Target score ≥4.5 average for excellence. All criteria ≥4."
+    },
+    "minimum_thresholds": {
+      "synthesis": "For synthesis work: Source Coverage ≥3 (must include relevant sources), Thematic Analysis ≥3 (must identify patterns), Conflict Resolution ≥3 if conflicts exist (must address, not ignore), Evidence & Citation ≥3 (must cite sources).",
+      "analogy": "For analogy work: Structural Mapping ≥3 (must have explicit mapping), Analogy Depth ≥3 (must be relational not just surface), Limitations ≥3 (MUST acknowledge where it breaks - critical safety check), Evidence & Citation ≥3 (must ground analogy).",
+      "overall_average": "Must be ≥3.5 across applicable criteria before delivering. Higher threshold (≥4.0) for complex or high-stakes work."
+    }
+  },
+  "common_failure_modes": {
+    "cherry_picking_sources": "Source Coverage: 1-2. Only includes sources supporting pre-existing view, ignores disconfirming evidence. Fix: Systematically search for disconfirming evidence. Include and address it explicitly.",
+    "single_source_pattern": "Thematic Analysis: 1-2. Claims pattern from one source ('X always happens because Source A said so'). Fix: Patterns require 3+ sources. One source = theme or anecdote, not pattern.",
+    "false_synthesis": "Conflict Resolution: 1-2. Forces consensus where genuine disagreement exists. 'All sources agree...' when they don't. Fix: Acknowledge conflicts. Use meta-framework, scope distinction, or state uncertainty.",
+    "surface_analogy": "Analogy Depth: 1-2. Maps surface attributes ('both are blue', 'both process information') not relational structure. Fix: Map relationships (A→B in source corresponds to X→Y in target with same causal pattern).",
+    "no_explicit_mapping": "Structural Mapping: 1-2. Vague 'X is like Y' without stating what corresponds to what. Fix: Create explicit mapping table: Entity A ↔ Entity X, Relation (A→B) ↔ Relation (X→Y).",
+    "analogy_overextension": "Limitations: 1-2. Pushes analogy beyond where it holds. No acknowledgment of where it breaks down. Fix: Explicitly state: 'Analogy holds for [X, Y] but breaks for [A, B] because [reason].'",
+    "no_new_insights": "Thematic Analysis (synthesis) or Productivity (analogy): 1-2. Just summarizes existing sources or restates target in source terms. No value-add. Fix: Ask 'What do we learn from combining sources / mapping domains that wasn't obvious before?'",
+    "mixing_facts_interpretations": "Evidence & Citation: 2-3. Presents interpretations as if they're facts. 'The data proves...' (interpretation) vs 'The data shows...' (fact). Fix: Label clearly: 'Data/Source shows X [fact]. This suggests Y [interpretation].'",
+    "vague_transfer": "Productivity: 2-3. 'This analogy could help with X' (vague). Fix: Specific transfer: 'In source, solution is A. Mapping to target, this becomes B [specific]. Evidence it works: [reasoning].'",
+    "ignoring_conflicts": "Conflict Resolution: 1-2. Sources disagree but analysis ignores it. Presents unified view when sources actually conflict. Fix: Identify conflicts explicitly. Resolve via meta-framework or state uncertainty."
+  },
+  "self_check_questions": [
+    "Synthesis: Have I included all relevant sources or am I cherry-picking?",
+    "Synthesis: Are patterns supported by 3+ sources or am I overgeneralizing from one?",
+    "Synthesis: Have I addressed conflicts between sources or ignored them?",
+    "Synthesis: What new insights emerge from synthesis that weren't in individual sources?",
+    "Analogy: Is my mapping explicit (what corresponds to what) or vague?",
+    "Analogy: Am I mapping relational structure (A→B maps to X→Y) or just surface features ('both are round')?",
+    "Analogy: Do multiple interconnected relations map (systematic) or just one isolated fact?",
+    "Analogy: Does analogy generate new predictions about target domain?",
+    "Analogy: Have I explicitly stated where analogy breaks down (limitations)?",
+    "Both: Are major claims backed by evidence with source citations?",
+    "Both: Do I distinguish facts (what sources say) from interpretations (my analysis)?",
+    "Both: Is this useful and actionable for the audience?",
+    "Both: Am I overconfident (claiming more than evidence supports)?",
+    "Overall: Would a skeptical expert in the domain accept this synthesis/analogy?"
+  ],
+  "evaluation_notes": "Synthesis & Analogy quality assessed across 8 weighted criteria. For synthesis: Source Coverage (1.3x), Thematic Analysis (1.4x), and Conflict Resolution (1.2x) are critical. For analogy: Structural Mapping (1.5x), Analogy Depth (1.4x), Productivity (1.3x), and Limitations (1.2x) are critical. Evidence & Citation (1.1x) applies to both. Criteria marked N/A for irrelevant task types (e.g., Structural Mapping N/A for synthesis-only). Minimum standard: ≥3.5 average across applicable criteria, with all criteria ≥3 individually. Higher threshold (≥4.0) for complex work. Limitations acknowledgment is CRITICAL for analogies to prevent dangerous overextension."
+}

skills/synthesis-and-analogy/resources/methodology.md

@@ -0,0 +1,331 @@
+# Advanced Synthesis & Analogy Methodology
+
+## Workflow
+
+```
+Advanced Synthesis & Analogy Progress:
+- [ ] Step 1: Advanced synthesis techniques for complex sources
+- [ ] Step 2: Structural mapping theory for deep analogies
+- [ ] Step 3: Cross-domain problem-solving and creative ideation
+- [ ] Step 4: Validity testing and refinement
+- [ ] Step 5: Integration and advanced applications
+```
+
+**Step 1**: Advanced synthesis - meta-synthesis, mixed methods, temporal synthesis. See [1. Advanced Synthesis](#1-advanced-synthesis-techniques).
+
+**Step 2**: Structural mapping - systematic correspondence, relational structure. See [2. Structural Mapping Theory](#2-structural-mapping-theory-deep-dive).
+
+**Step 3**: Creative problem-solving - analogical transfer, forced connections. See [3. Analogical Problem-Solving](#3-analogical-problem-solving).
+
+**Step 4**: Validity testing - systematicity, productivity, scope limits. See [4. Testing Analogy Validity](#4-testing-analogy-validity).
+
+**Step 5**: Integration - combining synthesis + analogy, avoiding pitfalls. See [5. Common Pitfalls](#5-common-pitfalls--how-to-avoid).
+
+---
+
+## 1. Advanced Synthesis Techniques
+
+### Meta-Synthesis (Synthesis of Syntheses)
+
+**When to use**: Combining multiple literature reviews or systematic reviews into higher-level synthesis.
+
+**Process:**
+1. **Gather existing syntheses**: Find 3-10 literature reviews, systematic reviews, or meta-analyses
+2. **Extract findings**: For each synthesis, note main conclusions, themes, effect sizes
+3. **Compare**: Where do syntheses agree? Disagree? What explains differences (population, methodology, timeframe)?
+4. **Higher-level patterns**: What emerges across all syntheses that wasn't obvious in any single one?
+5. **Synthesize**: Create narrative connecting all synthesis-level findings
+
+**Example**: Meta-synthesis of 5 systematic reviews on "effective onboarding" → Review A (tech): automation reduces friction. Review B (retail): personalization increases engagement. Review C (healthcare): compliance training critical. **Meta-synthesis**: Effective onboarding = (1) reduce friction (automation/simplification), (2) personalize to role/individual, (3) ensure compliance for regulated industries. All three necessary, priority varies by domain.
+
+### Mixed-Methods Integration (Qualitative + Quantitative)
+
+**Challenge**: Combining narrative/thematic data (interviews, observations) with numerical data (surveys, metrics).
+
+**Approaches:**
+
+| Approach | Description | When to Use | Example |
+|----------|-------------|-------------|---------|
+| **Convergent** | Collect qual + quant simultaneously, merge during analysis | Validate findings across methods | Survey shows 70% churn at Step 3 (quant). Interviews reveal "Step 3 too complex" (qual). **Integration**: Step 3 complexity causes 70% churn. |
+| **Explanatory** | Quant first (identify pattern), qual second (explain why) | Unexpected quant results need explanation | Metric: Feature X has low usage (quant). Interviews: "didn't know it existed" (qual). **Explanation**: Low usage due to discoverability, not value. |
+| **Exploratory** | Qual first (generate hypotheses), quant second (test at scale) | New domain, need to develop measures | Interviews identify 4 user personas (qual). Survey confirms distribution: 40%/30%/20%/10% (quant). **Integration**: Validated persona model with prevalence data. |
+| **Embedded** | One method primary, other secondary/supporting | One method dominant, other adds context | Experiment shows Feature A outperforms B (quant primary). User comments explain why (qual secondary). |
+
+### Temporal Synthesis (Longitudinal Integration)
+
+**When to use**: Synthesizing data points across time to understand trends, cycles, evolution.
+
+**Techniques:**
+- **Trend analysis**: Identify directional changes (increasing, decreasing, stable)
+- **Cycle detection**: Look for repeating patterns (seasonal, periodic)
+- **Event correlation**: Link events to outcome changes (did X cause shift in Y?)
+- **Phase transitions**: Identify inflection points where system shifts regimes
+
+**Example**: Synthesizing 24 months of customer satisfaction data → Month 1-8: declining (65→55%), trigger: pricing change in Month 2. Month 9-16: stable ~55%. Month 17-24: improving (55→70%), trigger: new onboarding in Month 17. **Synthesis**: Pricing hurt satisfaction short-term (-10pp), plateaued, new onboarding recovered and exceeded baseline (+5pp net).
+
+### Cross-Cultural Synthesis
+
+**Challenge**: Synthesizing findings across different cultural contexts where same phenomenon manifests differently.
+
+**Approach:**
+1. **Etic analysis**: Identify universal patterns (what's same across all cultures?)
+2. **Emic analysis**: Identify culture-specific manifestations (what's unique per culture?)
+3. **Synthesis**: "Universal pattern X manifests as A in Culture 1, B in Culture 2, C in Culture 3"
+
+**Example**: Synthesizing user research across US, Japan, Brazil on "collaboration tools" → **Etic (universal)**: All cultures value real-time communication and file sharing. **Emic (specific)**: US prefers async text (Slack), Japan prefers video (face-to-face culture), Brazil prefers voice (WhatsApp voice notes). **Synthesis**: Build core real-time communication + file sharing (universal), with mode preferences (cultural adaptation).
+
+---
+
+## 2. Structural Mapping Theory Deep Dive
+
+### Gentner's Structure-Mapping Theory
+
+**Core principle**: Good analogies map relational structure, not object attributes.
+
+**Components:**
+- **Entities/Objects**: Elements in domains (can differ)
+- **Attributes**: Properties of entities (can differ)
+- **Relations**: Connections between entities (**must correspond**)
+- **Higher-order relations**: Relations between relations (**strongest mapping**)
+
+**Example:**
+
+| Source: Solar System | Target: Atom |
+|---------------------|--------------|
+| **Entities**: Sun, planets | **Entities**: Nucleus, electrons (DIFFERENT objects) |
+| **Attributes**: Sun is hot, large | **Attributes**: Nucleus is positive charge (DIFFERENT properties) |
+| **Relations**: Planets orbit sun (mass attracts) | **Relations**: Electrons orbit nucleus (charge attracts) (**SAME structure**) |
+| **Higher-order**: More massive planet → elliptical orbit | **Higher-order**: Different energy electrons → different orbital shells (**SAME pattern**) |
+
+**Verdict**: Strong analogy because relational structure maps despite different entities/attributes.
+
+### Systematicity Principle
+
+**Definition**: Analogies are stronger when they map interconnected systems of relations, not isolated facts.
+
+**Test**: Does analogy map single relation or network of relations?
+
+**Weak (single relation)**: "Brain is like computer: both process information" (1 relation, vague)
+
+**Strong (systematic)**: "Brain is like computer:
+- Input devices (sensors) → Processing (neurons/CPU) → Output (motor/display) [information flow]
+- Storage (synaptic weights/RAM) ↔ Processing [interaction]
+- Feedback loops (learning/software updates) [adaptation]
+- **Systematic**: Maps entire input-processing-output-storage-feedback network"
+
+### Pragmatic Centrality
+
+**Principle**: Focus mapping on aspects most relevant to current goal, even if other mappings exist.
+
+**Example**: "Cell is like factory"
+- **Goal: Explain protein synthesis** → Map nucleus (blueprint storage) to office, ribosomes (assembly) to assembly line, mitochondria (power) to power plant. **Central**: Manufacturing process.
+- **Goal: Explain energy management** → Map mitochondria (ATP production) to power plant, glucose transport to fuel delivery. **Central**: Energy system.
+
+**Same analogy, different pragmatic focus based on goal.**
+
+---
+
+## 3. Analogical Problem-Solving
+
+### Retrieve-Map-Transfer Process
+
+**Step 1: Retrieve** - Find source domain with similar structural problem
+- **Spontaneous retrieval**: Reminded of similar problem from past experience
+- **Deliberate search**: Systematically search analogous domains
+- **Trick**: Look for structural similarity (relationships), not surface (objects)
+
+**Step 2: Map** - Align elements between source and target
+- Map entities (what corresponds to what?)
+- Map relations (A→B in source corresponds to X→Y in target?)
+- Test mapping validity
+
+**Step 3: Transfer** - Bring solution from source to target
+- Identify solution element in source
+- Map solution to target domain
+- Adapt as needed (exact transfer rare)
+
+**Example:**
+
+**Target problem**: Users abandoning mid-signup (structural issue: high friction in multi-step process)
+
+**Step 1: Retrieve source domain**: Restaurant ordering process (similar structure: multi-step, high abandonment at payment)
+
+**Step 2: Map**:
+- Source: Menu → Order → Customize → Pay (steps)
+- Target: Email → Password → Profile → Verify (steps)
+- **Both**: Multi-step funnel with dropoff at customization/profile (high cognitive load)
+
+**Step 3: Transfer solution**:
+- **Source solution**: Restaurants use "defaults" (chef's recommendation), allow skip customization, pay at end
+- **Target application**: Use defaults (auto-generate username), allow skip profile (fill later), verify at end (not middle)
+- **Result**: Reduced abandonment 40% by reordering steps and adding defaults
+
+### Forced Connections (Systematic Variation)
+
+**When to use**: Creative ideation, generating novel solutions.
+
+**Technique**: Systematically pair target problem with random domains, force analogy, see what emerges.
+
+**Process:**
+1. **State problem**: [What are you trying to solve?]
+2. **Random domain**: Pick unrelated domain (use random generator: biology, music, sports, architecture, cooking, weather...)
+3. **Force mapping**: "How is [problem] like [random domain]?" - Map structure even if feels forced
+4. **Extract**: What insights emerge from forced mapping?
+5. **Repeat**: Try 5-10 random domains
+
+**Example:**
+
+**Problem**: Improve code review process
+
+**Random domain 1: Restaurant food critique**
+- Map: Code → Dish, Reviewer → Food critic, Review → Critique
+- Insight: Critics use multi-criteria rubrics (presentation, taste, technique). **Transfer**: Create multi-criteria code review rubric (readability, correctness, architecture).
+
+**Random domain 2: Airport security**
+- Map: Code submission → Passenger, Review → Security check, Approval → Boarding
+- Insight: Tiered security (pre-check for trusted). **Transfer**: Trusted developers get lighter review, new contributors get thorough review.
+
+**Random domain 3: Jazz improvisation**
+- Map: Code changes → Musical variations, Codebase → Jazz standard, Review → Band synchronization
+- Insight: Jazz uses "call and response", tight feedback loops. **Transfer**: Pair programming as real-time "call and response" review.
+
+**Result**: 3 novel ideas from forced analogies (rubric-based review, tiered trust, pair programming emphasis).
+
+---
+
+## 4. Testing Analogy Validity
+
+### Systematicity Test
+
+**Question**: Does analogy map interconnected system of relations or just one fact?
+
+**Strong analogy**: Maps multiple interconnected relations
+**Weak analogy**: Maps single isolated fact
+
+**Test procedure:**
+1. List all mapped relations
+2. Check if relations interconnect (does Relation A affect Relation B?)
+3. Count: 3+ interconnected relations = systematic
+
+### Productivity Test
+
+**Question**: Does analogy generate new predictions or insights about target domain?
+
+**Productive analogy**: Leads to testable predictions in target
+**Unproductive analogy**: Just restates what we already know
+
+**Test procedure:**
+1. What does analogy predict about target that we didn't already know?
+2. Is prediction testable?
+3. If tested, does it hold?
+
+**Example**:
+- **Analogy**: "Immune system is like cybersecurity defense"
+- **Source domain fact**: Immune system has "memory" (antibodies persist)
+- **Prediction for target**: Cybersecurity should have "threat memory" (remember past attack signatures)
+- **Test**: Implement threat signature database. Does it improve detection? (Yes → productive analogy)
+
+### Scope Limitation Test
+
+**Question**: Where does analogy break down? What doesn't transfer?
+
+**Critical**: Every analogy has limits. Acknowledging them prevents overgeneralization.
+
+**Test procedure:**
+1. List source domain facts
+2. For each fact, does it map to target? (Yes/No)
+3. Explicitly state "No" cases as limitations
+
+**Example**:
+- **Analogy**: "Startup growth is like rocket launch"
+- **Maps**: Escape velocity (critical early momentum), fuel burn (runway), trajectory (growth curve)
+- **Breaks down**: Rockets don't pivot mid-flight, can't refuel in space, one-shot (not iterative)
+- **Limitation**: "Analogy useful for early momentum needs, but breaks down for pivot/iteration/long-term sustainability aspects. Startups ARE iterative and can refuel (fundraise), unlike rockets."
+
+---
+
+## 5. Common Pitfalls & How to Avoid
+
+| Pitfall | Description | How to Avoid |
+|---------|-------------|--------------|
+| **Surface Analogy** | Maps superficial features, not structure. "Both are blue" doesn't help. | Explicitly map relations, not attributes. Test: Does analogy work if you change objects? |
+| **False Synthesis** | Forcing agreement where genuine conflict exists. | When sources disagree, state it clearly. Don't paper over conflicts. Provide meta-framework or state uncertainty. |
+| **Cherry-Picking** | Selecting only sources/data that support pre-existing view. | Systematic source inclusion. Explicitly address disconfirming evidence. |
+| **Analogy as Proof** | Treating analogy as evidence rather than illustration. | State clearly: "Analogies illustrate, don't prove." Use analogy for explanation, not argumentation. |
+| **Overgeneralization** | One data point or source → sweeping pattern claim. | Require 3+ sources for pattern claim. Acknowledge n=1 as anecdote, not trend. |
+| **Overextending Analogy** | Pushing analogy beyond where it holds. | Explicitly test scope limits. State where analogy breaks down. Know when to stop. |
+| **Mixing Levels** | Confusing data (facts) with interpretation (analysis). | Clearly label what's observation vs interpretation. "Data shows X (fact). This suggests Y (interpretation)." |
+| **Ignoring Disconfirming** | Dismissing sources that contradict synthesis. | Actively seek disconfirming evidence. Explain why disconfirming source doesn't invalidate (scope, quality) OR revise synthesis. |
+| **Vague Mapping** | Unclear what corresponds to what in analogy. | Create explicit mapping table. "In source, X. In target, Y. X↔Y because [relation]." |
+| **Single-Source Pattern** | Claiming pattern from one source. | Patterns require repetition across sources. One source = theme, not pattern. |
+
+### Synthesis Quality Checklist
+
+- [ ] All relevant sources included and documented
+- [ ] Source quality assessed (not all sources equal)
+- [ ] Themes identified with frequency counts (X/Y sources)
+- [ ] Agreements clearly stated
+- [ ] Conflicts explicitly addressed and resolved (or uncertainty stated)
+- [ ] Patterns identified with evidence from 3+ sources
+- [ ] New insights generated (synthesis adds value beyond summarizing)
+- [ ] Evidence cited for claims
+- [ ] Gaps and uncertainties acknowledged
+- [ ] Conclusions proportional to evidence (not over-claiming)
+
+### Analogy Quality Checklist
+
+- [ ] Source domain familiar to audience
+- [ ] Target domain clearly defined
+- [ ] Structural mapping explicit (entities and relations mapped)
+- [ ] Multiple relations map (systematicity)
+- [ ] Generates new predictions (productivity)
+- [ ] Limitations explicitly stated (where it breaks down)
+- [ ] Not used as proof (illustrative only)
+- [ ] Appropriate for goal (pragmatic centrality)
+- [ ] Deep (relational) not surface (attribute-based)
+- [ ] Doesn't overextend beyond valid scope
+
+---
+
+## 6. Advanced Integration Techniques
+
+### Synthesis → Analogy Pipeline
+
+**Use case**: Synthesize complex findings, then use analogy to make accessible.
+
+**Process:**
+1. **Synthesize** first: Identify patterns across sources
+2. **Find familiar domain**: What domain does audience know that has similar structure?
+3. **Map pattern**: Transfer synthesized pattern to familiar domain
+4. **Explain**: Use analogy to make synthesis accessible
+
+**Example**:
+- **Synthesis**: Analyzed 20 incident postmortems → Pattern: 80% involve config change + missing rollback.
+- **Familiar domain**: Home renovation
+- **Map**: Config change = renovation, Rollback = "undo" plan, Missing rollback = no plan B if renovation fails
+- **Analogy**: "System configs are like home renovations. You can change layout (config), but without 'undo' plan (rollback), failed renovation leaves you homeless (broken system). Always have rollback plan."
+- **Result**: Technical pattern accessible to non-technical stakeholders via home renovation analogy.
+
+### Multiple Analogies (Triangulation)
+
+**Use case**: Complex target needs multiple analogies, each illuminating different aspect.
+
+**Technique**: Use 2-3 analogies for same target, each mapping different facet.
+
+**Example - Explaining "Technical Debt":**
+
+**Analogy 1: Financial debt**
+- Maps: Borrowing (quick hacks) → Interest (maintenance cost) → Compounding (code harder to change)
+- **Illuminates**: Cost over time, eventually must pay back
+
+**Analogy 2: Tooth decay**
+- Maps: Skipping brushing (skipping quality) → Cavity (bug) → Root canal (rewrite)
+- **Illuminates**: Gradual degradation, early prevention cheaper than late fix
+
+**Analogy 3: Garden weeds**
+- Maps: Weeds (bad code) → Spread (contagion) → Harder to remove when established
+- **Illuminates**: Contagion aspect, importance of early removal
+
+**Together**: Three analogies illuminate cost accumulation (financial), gradual degradation (tooth), and contagion (weeds). More complete understanding than any single analogy.
+
+**Key Takeaway**: Advanced synthesis combines information rigorously (thematic analysis, conflict resolution, pattern identification). Advanced analogy maps relational structure systematically (structural mapping theory, systematicity, productivity testing). Together, they enable understanding complex systems, transferring knowledge across domains, and generating creative solutions.

skills/synthesis-and-analogy/resources/template.md

@@ -0,0 +1,333 @@
+# Synthesis & Analogy Template
+
+## Workflow
+
+Copy this checklist and track your progress:
+
+```
+Synthesis & Analogy Template Progress:
+- [ ] Step 1: Define goal and scope
+- [ ] Step 2: Gather and organize sources/domains
+- [ ] Step 3: Apply synthesis or analogy techniques
+- [ ] Step 4: Extract insights and test validity
+- [ ] Step 5: Document findings and validate quality
+```
+
+**Step 1**: Define goal - synthesis question or analogy purpose. See [Section 1](#1-goal--scope).
+
+**Step 2**: Gather sources/domains - list all sources (synthesis) or identify source/target domains (analogy). See [Section 2A](#2a-synthesis-sources) or [Section 2B](#2b-analogy-domains).
+
+**Step 3**: Apply techniques - thematic synthesis, conflict resolution (synthesis) or structural mapping (analogy). See [Section 3A](#3a-synthesis-application) or [Section 3B](#3b-analogy-application).
+
+**Step 4**: Extract insights - patterns, themes, transferred knowledge. See [Section 4](#4-insights--findings).
+
+**Step 5**: Validate quality - use quality checklist before finalizing. See [Quality Checklist](#quality-checklist).
+
+---
+
+## 1. Goal & Scope
+
+**Choose One** (or combine):
+- [ ] **Synthesis**: Combining multiple sources into unified understanding
+- [ ] **Analogy**: Transferring knowledge between domains or explaining via familiar concepts
+- [ ] **Both**: Synthesize then use analogy to explain
+
+**Goal Statement:**
+
+[What question are you answering? What problem are you solving? What are you trying to understand or explain?]
+
+**Success Criteria:**
+
+What does good output look like?
+- [ ] [Criterion 1 - e.g., "Identifies 3-5 major themes across all sources"]
+- [ ] [Criterion 2 - e.g., "Resolves apparent conflicts between sources"]
+- [ ] [Criterion 3 - e.g., "Generates actionable insights not present in individual sources"]
+
+---
+
+## 2A. Synthesis: Sources
+
+*Skip this section if doing analogy only.*
+
+### Source Inventory
+
+List all sources to synthesize:
+
+| Source # | Type | Title/Description | Key Claims | Date | Quality |
+|----------|------|-------------------|------------|------|---------|
+| 1 | [Research paper / Interview / Survey / Data / Report] | [Title or description] | [Main points] | [When] | [High/Med/Low] |
+| 2 | [Type] | [Title] | [Points] | [When] | [Quality] |
+| 3 | [Type] | [Title] | [Points] | [When] | [Quality] |
+| 4 | [Type] | [Title] | [Points] | [When] | [Quality] |
+| 5 | [Type] | [Title] | [Points] | [When] | [Quality] |
+
+**Total sources**: [Count]
+
+**Source quality notes**: [Any concerns about reliability, bias, or relevance of sources?]
+
+---
+
+## 2B. Analogy: Domains
+
+*Skip this section if doing synthesis only.*
+
+### Source Domain (Familiar)
+
+**Domain**: [What domain are we drawing from? Must be familiar to audience.]
+
+**Why this domain?** [Why is this a good source for analogy? What makes it familiar and well-understood?]
+
+**Key entities in source domain:**
+- Entity 1: [Name and description]
+- Entity 2: [Name and description]
+- Entity 3: [Name and description]
+
+**Key relationships in source domain:**
+- Relationship 1: [How entities relate - e.g., "Entity 1 → Entity 2 (causes, contains, transforms)"]
+- Relationship 2: [Relationship description]
+- Relationship 3: [Relationship description]
+
+### Target Domain (Unfamiliar or Complex)
+
+**Domain**: [What are we trying to explain or solve?]
+
+**Why needs explanation/transfer?** [What's complex or unfamiliar about this domain?]
+
+**Key entities in target domain:**
+- Entity A: [Name and description]
+- Entity B: [Name and description]
+- Entity C: [Name and description]
+
+**Key relationships in target domain:**
+- Relationship A: [How entities relate]
+- Relationship B: [Relationship description]
+- Relationship C: [Relationship description]
+
+---
+
+## 3A. Synthesis: Application
+
+*Use if synthesizing multiple sources.*
+
+### Thematic Analysis
+
+**Theme 1: [Theme Name]**
+- **Description**: [What is this theme about?]
+- **Frequency**: [X/Y sources mention this]
+- **Supporting sources**: [List source #s]
+- **Key evidence**:
+  - Source [#]: "[Quote or key point]"
+  - Source [#]: "[Quote or key point]"
+  - Source [#]: "[Quote or key point]"
+- **Importance**: [Why does this matter? What's the implication?]
+
+**Theme 2: [Theme Name]**
+[Same structure as Theme 1]
+
+**Theme 3: [Theme Name]**
+[Same structure as Theme 1]
+
+**Theme 4: [Theme Name]** (if applicable)
+[Same structure as Theme 1]
+
+**Theme 5: [Theme Name]** (if applicable)
+[Same structure as Theme 1]
+
+### Agreements Across Sources
+
+**What do most/all sources agree on?**
+
+| Agreement | Sources Supporting | Evidence |
+|-----------|-------------------|----------|
+| [Point of agreement] | [Source #s] | [Brief supporting quotes/data] |
+| [Point of agreement] | [Source #s] | [Evidence] |
+| [Point of agreement] | [Source #s] | [Evidence] |
+
+### Conflicts & Resolution
+
+**Conflict 1:**
+- **Source(s) A claim**: [What do some sources say?] (Sources: [#s])
+- **Source(s) B claim**: [What do other sources say?] (Sources: [#s])
+- **Nature of conflict**: [Genuine disagreement or scope/context difference?]
+- **Resolution**: [How can we reconcile? Meta-framework? Scope distinction? Temporal? State uncertainty?]
+
+**Conflict 2:**
+[Same structure as Conflict 1]
+
+**Conflict 3:**
+[Same structure as Conflict 1]
+
+### Patterns & Meta-Insights
+
+**Pattern 1:**
+- **Pattern**: [What repeats across multiple sources in different guises?]
+- **Evidence**: [Where do we see this? Source #s and examples]
+- **Meta-insight**: [What does this pattern tell us? What's the underlying principle?]
+
+**Pattern 2:**
+[Same structure as Pattern 1]
+
+**Gaps & Uncertainties:**
+
+What's not covered or unclear?
+- Gap 1: [What's missing from all sources?]
+- Gap 2: [What remains uncertain despite synthesis?]
+- Gap 3: [What contradictions remain unresolved?]
+
+---
+
+## 3B. Analogy: Application
+
+*Use if creating or testing analogy.*
+
+### Structural Mapping Table
+
+| Source Domain (Familiar) | → | Target Domain (Unfamiliar) | Validity |
+|--------------------------|---|----------------------------|----------|
+| **Entity Mapping** | | | |
+| [Source entity 1] | ↔ | [Target entity A] | [Does this map hold? Y/N, why?] |
+| [Source entity 2] | ↔ | [Target entity B] | [Validity] |
+| [Source entity 3] | ↔ | [Target entity C] | [Validity] |
+| **Relationship Mapping** | | | |
+| [Source relationship: X→Y] | ↔ | [Target relationship: A→B] | [Do similar causal/structural relations exist?] |
+| [Source relationship] | ↔ | [Target relationship] | [Validity] |
+| [Source relationship] | ↔ | [Target relationship] | [Validity] |
+
+### Deep vs Surface Analysis
+
+**Deep (Structural) Similarities:**
+- [What structural/relational similarities exist? These make the analogy powerful.]
+- [Example: "Both have feedback loops", "Both involve hub-spoke topology", "Both use hierarchical organization"]
+
+**Surface (Superficial) Similarities:**
+- [What surface-level similarities exist but don't help understanding?]
+- [Example: "Both are round", "Both involve numbers" - note these are weak and should not drive analogy]
+
+**Assessment**: [Is this analogy primarily deep (good) or surface (weak)?]
+
+### Transfer of Insights
+
+**What transfers from source to target?**
+
+1. **Insight/Solution 1**: [What knowledge from source domain applies to target?]
+   - Source: [How it works in source domain]
+   - Target: [How it could work in target domain]
+   - Evidence it transfers: [Why does this mapping work?]
+
+2. **Insight/Solution 2**:
+   [Same structure as Insight 1]
+
+3. **Insight/Solution 3**:
+   [Same structure as Insight 1]
+
+### Limitations & Where Analogy Breaks Down
+
+**Critical**: Always acknowledge where analogy stops working.
+
+**Limitation 1**:
+- **What doesn't transfer**: [Aspect of source domain that doesn't map to target]
+- **Why it breaks**: [Explanation]
+- **Implication**: [What this means for using the analogy]
+
+**Limitation 2**:
+[Same structure as Limitation 1]
+
+**Limitation 3**:
+[Same structure as Limitation 1]
+
+**Bottom line**: [This analogy is useful for understanding [X, Y, Z] but should not be pushed to explain [A, B, C].]
+
+---
+
+## 4. Insights & Findings
+
+### Synthesized Summary
+
+**Main findings** (3-5 bullet points):
+- [Finding 1 from synthesis/analogy work]
+- [Finding 2]
+- [Finding 3]
+- [Finding 4]
+- [Finding 5]
+
+**New insights not present in individual sources:**
+
+[What have we learned from combining sources or mapping analogies that wasn't obvious from any single source? This is the value-add of synthesis/analogy work.]
+
+### Supporting Evidence
+
+**Claim 1**: [Major claim or insight]
+- **Evidence**: [Which sources support this? Specific quotes or data points]
+- **Strength**: [How strong is the evidence? High/Medium/Low]
+
+**Claim 2**: [Major claim or insight]
+[Same structure as Claim 1]
+
+**Claim 3**: [Major claim or insight]
+[Same structure as Claim 1]
+
+### Actionable Implications
+
+**What should we do based on these insights?**
+
+1. **Action 1**: [Specific action]
+   - **Rationale**: [Why this action follows from synthesis/analogy]
+   - **Priority**: [High/Medium/Low]
+   - **Owner**: [Who should do this]
+
+2. **Action 2**:
+   [Same structure as Action 1]
+
+3. **Action 3**:
+   [Same structure as Action 1]
+
+### Confidence & Uncertainties
+
+**High confidence in:**
+- [What are we very sure about based on synthesis/analogy?]
+
+**Medium confidence in:**
+- [What seems likely but has some uncertainty?]
+
+**Low confidence / Open questions:**
+- [What remains uncertain or needs further investigation?]
+
+---
+
+## Quality Checklist
+
+Before finalizing, verify:
+
+**For Synthesis:**
+- [ ] All relevant sources included in inventory (no cherry-picking)
+- [ ] Thematic analysis identifies 3-5 major themes with supporting evidence
+- [ ] Agreements across sources documented (what's consensus?)
+- [ ] Conflicts explicitly addressed and resolved (or stated as uncertain)
+- [ ] Patterns identified beyond what individual sources state (value-add)
+- [ ] Evidence cited for major claims (source #s, quotes, data)
+- [ ] Gaps and uncertainties acknowledged (what's missing or unclear?)
+- [ ] New insights generated (synthesis adds value, not just summarizes)
+- [ ] Facts distinguished from interpretations (clear what's data vs analysis)
+
+**For Analogy:**
+- [ ] Source domain appropriate (familiar to audience, well-understood)
+- [ ] Target domain clear (what we're explaining or transferring to)
+- [ ] Structural mapping table complete (entities and relationships mapped)
+- [ ] Deep (structural) similarities identified (not just surface features)
+- [ ] Mapping validity tested (do relationships actually transfer?)
+- [ ] Insights/solutions transfer clearly explained (what moves from source to target)
+- [ ] Limitations explicitly stated (where analogy breaks down)
+- [ ] Analogy doesn't overextend (knows when to stop)
+- [ ] Helps understanding (audience will grasp target domain better)
+- [ ] Not used as proof (analogies illustrate, don't prove)
+
+**Both:**
+- [ ] Goal from Step 1 achieved (answered question or solved problem)
+- [ ] Success criteria from Step 1 met (check against your own criteria)
+- [ ] Evidence-based (grounded in sources/domains, not speculation)
+- [ ] Actionable implications provided (so what? what should we do?)
+- [ ] Appropriate level of confidence stated (not overconfident or too hedged)
+- [ ] Readable and clear (audience will understand)
+- [ ] No contradictions (internally consistent)
+
+**Minimum Standard**: All applicable checklist items should be checkable. Average rubric score ≥ 3.5/5.