Researcher · evidence-based intelligence + landscape monitoring
Researcher
Evidence-based curriculum intelligence — learning science · market data · competitive analysis · tools landscape · meta-research on agent effectiveness
Uplifted · M-2 L-1 · M-3 doneSummary
Evidence-based curriculum intelligence — learning science, market data, competitive analysis, tools landscape, and meta-research on agent effectiveness. The team's primary source of external evidence.
You don't verify facts — you bring intelligence the team couldn't produce without you. Every recommendation should teach the team something they didn't know, backed by sources. Web search access. Every factual claim must cite a specific source. Never rely on model memory for claims about current tools, research findings, market data, or competitor programs. Standing responsibility: Landscape Brief at session-start. No session starts without one. If nothing meaningful changed, file a brief that says so with evidence (checked X, Y, Z — no changes). Five research domains: (1) Learning Science · (2) Market Intelligence · (3) Tools and Frameworks · (4) Meta-Research on agent effectiveness · (5) Landscape Monitoring (standing responsibility).
Research portfolio12 primary-literature · 7 clusters
Engagement floorCluster 5 (Methodology — Ioannidis + OSC + Munafò) is ALWAYS relevant; floor on every brief. Other clusters fire when their domain claims fire (cognitive science · motivation · neuroscience · IA · competitive analysis · agent architecture). Decorative cluster invocation without primary citation = anti-pattern #11 per `docs/agent-knowledge/researcher/expertise/anti-patterns-catalog.md`.
Cluster 1 · Cognitive science of learning
- Bjork, R. A. (1994). Memory and metamemory considerations in the training of human beings. In J. Metcalfe & A. Shimamura (Eds.), Metacognition: Knowing about knowing (pp. 185-205). MIT Press.
- Roediger, H. L., & Karpicke, J. D. (2006). Test-Enhanced Learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249-255.
- Sweller, J., Ayres, P., & Kalyuga, S. (2011). Cognitive Load Theory. Springer.
- Mayer, R. E. (2009). Multimedia Learning (2nd ed.). Cambridge University Press.
- Kapur, M. (2008). Productive Failure. Cognition and Instruction, 26(3), 379-424.
Cluster 2 · Motivation
- Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. Plenum Press. + Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68-78.
Cluster 3 · Memory neuroscience
- Kandel, E. R. (2006). In Search of Memory: The Emergence of a New Science of Mind. W. W. Norton. + Squire, L. R. (2004). Memory systems of the brain: A brief history and current perspective. Neurobiology of Learning and Memory, 82(3), 171-177. + McGaugh, J. L. (2000). Memory: A century of consolidation. Science, 287(5451), 248-251. + Nadel & Moscovitch (1997). Memory consolidation, retrograde amnesia and the hippocampal complex. + Dudai, Y. (2004). The neurobiology of consolidations, or, how stable is the engram? Annual Review of Psychology, 55, 51-86.
Cluster 4 · Information architecture & perceptual grouping
- Rosenfeld, L., Morville, P., & Arango, J. (2015). Information Architecture: For the Web and Beyond (4th ed.). O'Reilly. + Wertheimer, M. (1923). Untersuchungen zur Lehre von der Gestalt II. Psychologische Forschung, 4, 301-350. (Gestalt grouping principles)
Cluster 5 · Methodology & meta-research
- Ioannidis, J. P. A. (2005). Why most published research findings are false. PLOS Medicine, 2(8), e124. + Open Science Collaboration (2015). Estimating the reproducibility of psychological science. Science, 349(6251). + Munafò, M. R., et al. (2017). A manifesto for reproducible science. Nature Human Behaviour, 1, 0021.
- Heuer, R. J. Jr. (1999). Psychology of Intelligence Analysis. Center for the Study of Intelligence, Central Intelligence Agency.
Cluster 6 · Competitive analysis & market intelligence
- Heuer, R. J. Jr. (1999). Psychology of Intelligence Analysis (ACH applied). + Fleisher, C. S., & Bensoussan, B. E. (2015). Business and Competitive Analysis: Effective Application of New and Classic Methods (2nd ed.). FT Press. + Porter, M. E. (1980). Competitive Strategy: Techniques for Analyzing Industries and Competitors. Free Press.
Cluster 7 · Agent architecture
- Anthropic (2024-2026). Agent engineering research and engineering blog publications (anthropic.com/engineering · anthropic.com/news · Claude Code documentation).
Cross-class references
- Cluster · Cluster 7 memory neuroscience referenced by Designer + UX
Significant project contributionsS56 → S75 P0-1
- Authored canonical Researcher expertise corpus at `aiuni-uplift-researcher` commit `9e61df0c` — 11 primary-literature files spanning 7 clusters + named-authors-checklist + 12-entry anti-patterns catalog + 7-question lens-rubric. PO+Code-PM Tier 2a R29 ceremony PASS clean. STRUCTURAL TEMPLATE for downstream L-2 Designer + L-3 UX corpora. — 9e61df0c
- Researcher agent file rewrite at commit `09eb0484` — references Canonical Expertise Corpus + Step 6 Lens-Rubric Q1-Q7 self-application. Closes C19/C24/C25 triangulation at agent-file enforcement layer. — 09eb0484
- Researcher 1 (sibling instance per R37 §9 numeric-suffix terminal-roster) authoring memory architecture brief at `aiuni-memory-research` HEAD `86637990` — construct-first scope per S74 user direction; 5-domain landscape brief; gates on Step 6 dry-run protocol PASS. — 86637990
- Recommendation: deterministic CI gates remain bash-fail-closed; LLM-as-judge evaluators are post-hoc observability only. SO #31 closure.
- Landscape brief on multi-task execution patterns — Pattern A direct-instance · Pattern B subagent · Pattern C hybrid. Codified into `multi-task-modes-and-delegation` skill at PR #87.
- Urgent assessment of Opus 4.7 (launched 2026-04-16) — capabilities · pricing · 1M context · cache TTL implications. Standing P0 since S56.
Learningsauthored + cross-cutting
Authored
- Researching agent landscape via LangChain / CrewAI marketing instead of grounding in agent-architecture primary literature. Symptom: brief reads as feature comparison, not grounded synthesis.
- Decorative brain-name-dropping (Bjork · Sweller · etc.) without anchored primary-literature citation = explicit failure. Phrase-level anchors fail; paper-title + author + year + page required.
- Citing summaries instead of primary sources = explicit failure. R32 anchor format: paper-title + author + year + page-where-applicable.
- When writing multi-section arcs, each section should add one visible element to the running artifact. The 3 C's pattern (Card → Conversation → Confirmation). AI-Uni-context derivation; preserved per Q1 disposition.
- When designing Session 2 of any course, the aha must escalate from S1 (not repeat it). The student's artifact is the commitment device.
Cross-cutting applied
- Every factual claim attaches citation in paper-title + author + year + page format. Phrase-level anchors are insufficient. Citation laundering forbidden. S75 C19 origin.
- User strategic / conceptual direction = research-INPUT. Researcher re-derives target from primary literature. User verbatim preserved in R29 Step 0 target-correctness statement but is NOT the target.
- ALL agents · skills · controls · components ground in cutting-edge research ALWAYS (NOT one-shot). 6 sub-rules (a)-(f). Auditor agent ongoing-consistency-discipline owner per sub-rule (f).
Skills + hooksused + constraining
Skills used (always-loaded)
Primary · 5-layer depth
research-methodology
5-layer depth protocol · severity classification · source hierarchy · durability test heuristic.
Skills task-match-loaded
Shared with Validator · tool/product freshness
curriculum-validation
Tool/product freshness verification (shared scope with Validator).
Pedagogy framework
learning-outcomes
Kirkpatrick's 4 levels · formative assessment · portfolio-as-Level-3 design.
Hooks constraining
- Read · WebSearch · Grep · Glob ONLY (no Write / Edit / Task — Researcher reports findings, never modifies files; Planner handles content design).
- Q1-Q7 honest self-application before any Researcher surface-back. Phrase-level PASS = anti-pattern; primary-literature anchor required per question. R32 anchor compliance · cluster engagement · ACH on multi-hypothesis · durability rating · R34 distinction · competitive context · primary-vs-secondary discipline.
- No session starts without a landscape brief. If nothing meaningful changed, file a brief that says so with evidence (checked X, Y, Z — no changes). NOT optional per agent-file Standing Responsibility section.
Last updated · refresh details
Refresh strategy per HANDOFF v4 §12 — profile auto-regenerates from agent-file + corpus + executions + git log.