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 done
  • Agent file.claude/agents/researcher.md @ 09eb0484
  • Expertise corpusdocs/agent-knowledge/researcher/expertise/ @ 9e61df0c
  • Team membershipUplift cohort · Specialized direct-instance roster
  • Direct-instance worktreeaiuni-uplift-researcher on s75-uplift-researcher
  • ToolsRead · WebSearch · Grep · Glob
  • Catalog sectionAGENT-ROLES-AND-RESPONSIBILITIES-CATALOG §3.7
  • Existence statusexists

Summary

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).

Distinguishing characteristicsDifferent from Planner because: Researcher provides evidence + citations, Planner sequences curriculum. Different from Validator because: Researcher brings external intelligence, Validator audits internal content. Per R34 (S75 C24): Researcher brief precedes Designer dispatch when user direction is strategic/conceptual.

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

Domain: retention · transfer · sequencing

  • 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.
    Desirable difficulties · spacing · interleaving · variability · retrieval · forgetting-as-friend-of-learning
    bjork-desirable-difficulties.md @ 9e61df0c
  • Roediger, H. L., & Karpicke, J. D. (2006). Test-Enhanced Learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249-255.
    Testing effect · retrieval practice · transfer-appropriate processing · meta-analysis effect sizes
    roediger-karpicke-testing-effect.md @ 9e61df0c
  • Sweller, J., Ayres, P., & Kalyuga, S. (2011). Cognitive Load Theory. Springer.
    CLT triad · intrinsic / extraneous / germane load · element interactivity · expertise reversal
    sweller-cognitive-load.md @ 9e61df0c
  • Mayer, R. E. (2009). Multimedia Learning (2nd ed.). Cambridge University Press.
    Twelve principles · coherence · signaling · spatial-contiguity · personalization · embodiment
    mayer-multimedia-learning.md @ 9e61df0c
  • Kapur, M. (2008). Productive Failure. Cognition and Instruction, 26(3), 379-424.
    Productive failure · invention learning · designed-difficulty · activation-then-instruction sequence
    kapur-productive-failure.md @ 9e61df0c

Cluster 2 · Motivation

Domain: autonomy · competence · relatedness

  • 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.
    Self-determination theory (SDT) · three psychological needs · autonomy / competence / relatedness · intrinsic vs extrinsic motivation
    deci-ryan-self-determination.md @ 9e61df0c

Cluster 3 · Memory neuroscience

Domain: brain-level mechanism · consolidation

  • 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.
    Long-term potentiation (LTP) · consolidation · multiple-trace theory · reconsolidation · declarative vs procedural memory · hippocampal-cortical interaction
    memory-neuroscience-cluster.md @ 9e61df0c

Cluster 4 · Information architecture & perceptual grouping

Domain: content structure · navigation · labeling · Gestalt

  • 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)
    IA components — organization / labeling / navigation / search · Gestalt principles — proximity / similarity / closure / continuity / common-fate · IA-as-architecture-vs-content distinction
    rosenfeld-morville-information-architecture.md @ 9e61df0c

Cluster 5 · Methodology & meta-research

Domain: replication · reproducibility · ACH

  • 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.
    Replication crisis · p-hacking · pre-registration · meta-analysis · effect-size reporting · publication bias · always-relevant floor
    ioannidis-replication-and-meta-research.md @ 9e61df0c
  • Heuer, R. J. Jr. (1999). Psychology of Intelligence Analysis. Center for the Study of Intelligence, Central Intelligence Agency.
    Analysis of Competing Hypotheses (ACH) · ch. 8 evidence-matrix-and-disprove-leading · cognitive biases in analysis · structured analytic techniques
    heuer-ach-competitive-analysis.md @ 9e61df0c

Cluster 6 · Competitive analysis & market intelligence

Domain: competitor analysis · industry-structure

  • 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.
    ACH applied to competitor analysis · 24 business analysis methods · Five Forces · competitive positioning · industry structure
    heuer-ach-competitive-analysis.md @ 9e61df0c
    Note: Combined with Cluster 5 in corpus — Heuer ACH spans both methodology and competitive-analysis applications.

Cluster 7 · Agent architecture

Domain: agent-effectiveness · multi-agent design

  • Anthropic (2024-2026). Agent engineering research and engineering blog publications (anthropic.com/engineering · anthropic.com/news · Claude Code documentation).
    Agent architecture patterns · prompt design for agents · multi-agent collaboration · tool-use design · Claude Code harness patterns
    anthropic-agent-engineering-research.md @ 9e61df0c

Cross-class references

  • Cluster · Cluster 7 memory neuroscience referenced by Designer + UX
    Designer + UX cross-reference Cluster 7 rather than duplicating per Q4 disposition. Researcher corpus is canonical source.

Significant project contributionsS56 → S75 P0-1

  • S75 P0-1 · 2026-05-05 · L-1 cascade lane (substrate-foundation)
    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
  • S75 P0-1 · 2026-05-05 · M-3 codification
    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
  • S75 P0-1 (in-flight) · 2026-05-05 · Memory architecture brief (Researcher 1 sibling instance)
    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
  • S73 · 2026-05-02 · LangSmith assessment brief → SO #31
    Recommendation: deterministic CI gates remain bash-fail-closed; LLM-as-judge evaluators are post-hoc observability only. SO #31 closure.
  • S71 turn 14 · 2026-04-30 · Lane 23 multi-task delegation R1-R5 → PR #87
    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.
  • S56 · 2026-04-16 · Opus 4.7 urgent assessment
    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

  • Anti-pattern — vendor-survey-instead-of-first-principles
    Researching agent landscape via LangChain / CrewAI marketing instead of grounding in agent-architecture primary literature. Symptom: brief reads as feature comparison, not grounded synthesis.
  • Anti-pattern #11 — decorative cluster invocation without primary citation
    Decorative brain-name-dropping (Bjork · Sweller · etc.) without anchored primary-literature citation = explicit failure. Phrase-level anchors fail; paper-title + author + year + page required.
  • Citation laundering forbidden
    Citing summaries instead of primary sources = explicit failure. R32 anchor format: paper-title + author + year + page-where-applicable.
  • S43 origin — Progressive elicitation
    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.
  • S43 origin — Session 2 retention
    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

  • R32 anchor-grounding granularity
    Every factual claim attaches citation in paper-title + author + year + page format. Phrase-level anchors are insufficient. Citation laundering forbidden. S75 C19 origin.
  • R34 — research-input-vs-design-target distinction
    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.
  • Cutting-edge research as standing discipline
    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

  • Tool-scope · agent file frontmatter
    Read · WebSearch · Grep · Glob ONLY (no Write / Edit / Task — Researcher reports findings, never modifies files; Planner handles content design).
  • Lens-Rubric self-application (Step 6 codification)
    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.
  • Landscape Brief at session-start
    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

  • Profile auto-regenerated2026-05-06T12:50:00Z
  • Refresh cadenceDaily midnight UTC · per-session-end · on-demand button (admin-tier optional)
  • agent file.claude/agents/researcher.md @ 09eb0484
  • corpusdocs/agent-knowledge/researcher/expertise/ @ 9e61df0c
  • catalogdocs/agent-knowledge/AGENT-ROLES-AND-RESPONSIBILITIES-CATALOG.md @ d7d1aee8+241affb8

Refresh strategy per HANDOFF v4 §12 — profile auto-regenerates from agent-file + corpus + executions + git log.

Cross-references from the relationship graph

Edges sourced from cross-refs/matrix.json · also browsable on the org chart.

Cross-references 3 skills · 2 hooks

Skills · always-loaded

  • Research-Methodology
    5-layer depth · severity classification · source hierarchy · durability test.

Skills · task-match-loaded

  • Curriculum-Validation
    Tool/product freshness verification (shared with Validator).
  • Learning-Outcomes
    Kirkpatrick's 4 levels framework consumed for Level-3 portfolio design.

Hooks constraining

  • tool-scope-frontmatter · hard
    Read · WebSearch · Grep · Glob ONLY
  • lens-rubric-self-application · process-discipline
    Q1-Q7 before any brief surface-back