Section 9 · 23 cataloged artifacts · 11 pending
Research library.
Research artifacts authored across the team — Meta-Researcher landscape brief, per- agent expertise corpora, Memory Architecture brief, sibling landscape briefs. Source-of-truth: RESEARCH-ARTIFACT-INGESTION-CATALOG.md.
S75 · committed artifacts (13)
S75-1 · meta-researcher · track: cross-track
META-RESEARCHER-WORLD-CLASS-AGENTS-BRIEF-S75.md (M-1 brief)
M-1 brief: per-class corpus-of-expertise catalog; substrate-foundation for M-2..M-5 lanes. ~16K words / 116 references / 8 agent classes.
S75-2 · researcher · track: cross-track
Researcher Uplift L-1 corpus
M-2 L-1 expertise corpus for Researcher class. 11 primary-lit files + 4 standalone. Bjork · Roediger-Karpicke · Sweller · Mayer · Kapur · Deci-Ryan · memory-neuroscience-cluster · Rosenfeld-Morville-Wertheimer · Ioannidis · Heuer ACH · Anthropic agent-engineering.
S75-3 · designer · track: cross-track
Designer Uplift L-2 corpus
M-2 L-2 expertise corpus for Designer class. 9 primary-lit files + 4 standalone. Norman · Tognazzini · Krug · Müller-Brockmann · Sweller-applied · Mayer-applied · Hick+Fitts+Miller+Cowan · WCAG 2.2 · Frost atomic-design + Buxton.
S75-4 · po · track: prototype
KB-V4-VISION-CLARIFICATION-S75.md
2-layer architecture decomposition + v3.1 deploys reframe + R29 Step 0 self-applied. 2269 words / 194 lines.
S75-5 · po · track: cross-track
terminal-tracking-discipline.md (R37 + R38 NEW)
R37 NEW (canonical naming + open/close + track column) + R37 §9 + §10 + R38 NEW (PO-dispatch visibility-loop) + §13 HANDOFF_PACKET authoring template. ~415 LOC.
S75 UX L-3 corpus · ux · track: cross-track
UX Uplift L-3 corpus
M-2 L-3 expertise corpus for UX class. 10 primary-lit files + 4 standalone. 4 clusters — Cluster A Heuristic foundation (Nielsen + ISO 9241-110) · Cluster B Education-specific (Reeves-Reeves · Squires-Preece · Quinn) · Cluster C AI-specific (Amershi · Liao-Gruen-Miller · Microsoft 2023) · Cluster D Emotional/affective (Norman 2004 · Csikszentmihalyi 1990).
S75-7 · po · track: cross-track
AGENT-ROLES-AND-RESPONSIBILITIES-CATALOG.md
Comprehensive 12-entry per-agent catalog (10 existing + 2 proposed) with 9-field structure per agent + user-flagged confusion disambiguations + Tier 2 cascade snapshot.
S75-8 · po · track: cross-track
cutting-edge-research-as-standing-discipline.md
Cross-cutting principle: 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).
S75-9 · po · track: prototype
AGENT-DIRECTORY-WEB-APP-HANDOFF-S75.md (v4 corporate intranet)
AI Uni Corporate Intranet HANDOFF_PACKET — 11-section structure + 8-field bio + entitlement model + 7-step pipeline + refresh strategy + appendices. v1 MVP scope sections 1+3+4+5+8+9.
S75-10 · po · track: cross-track
cc-execution-discipline-and-po-orchestration-scope.md (R39)
R39 NEW: CC-execution discipline + PO orchestration scope. CF-4 PO-direct-write carve-out NARROWED. Pattern split codified.
S75-Stage1-mockup · designer + ux (dual-lead) · track: prototype
AI Uni Corporate Intranet Stage 1 mockup
Designer + UX dual-lead Stage 1 mockup deliverable — 7 HTML mockups + design-tokens.md + IA-flow.md + tokens.css 907 LOC + lens-rubric-self-application.md. WCAG 2.2 AA standing-gate · Lens-Rubric Q1-Q7 PASS WITH NOTES.
S75-Tester-L4 · tester · track: cross-track
Tester Uplift L-4 corpus
M-2 L-4 expertise corpus for Tester class. 11 primary-lit files + 4 standalone + 5 clusters (A discount-usability Nielsen-Landauer/Nielsen 1989/Krug 2009 · B think-aloud Ericsson-Simon · C blind-agentic Heriot-Watt 2025 + FeatureBench 2026 + Brookings 2025 + Anthropic 2026 · D mobile-first Wroblewski + Marcotte · E cognitive-bias Heuer 1999) + 13 anti-patterns + Q1-Q7 lens-rubric. Greenfield disposition (no pre-existing baseline).
S75-Tester-M3 · tester · track: cross-track
Tester agent file (.claude/agents/tester.md) — first M-3 greenfield
First M-3 agent-file authored from scratch (greenfield disposition; distinct from L-1/L-2/L-3 supersede-with-cross-reference pattern). References Canonical Expertise Corpus + 14-step engagement protocol + Q1-Q7 lens-rubric self-application before substantive output.
S75 · in-flight (1)
S75-6 · researcher 1 (sibling instance per R37 §9) · track: memory
MEMORY-ARCHITECTURE-RESEARCH-BRIEF-S75.md
Cognitive architecture research for designing OUR memory machine. 5-domain landscape brief — cognitive science · AI agent memory architectures · cross-session retrieval · compaction recovery · storage format trade-offs. Target 8-15K words.
S74 · committed artifacts (6)
S74-1 · Designer (KB strategic lane) · track: prototype
KB-DESIGN-RATIONALE-S74.md
Lane 0 BUILD CUSTOM verdict + Feed-Not-Hunt philosophy + FP-1..FP-5 + v3 anti-patterns; load-bearing for KB v4 design context
S74-2 · researcher · track: cross-track
Landscape brief: anthropic-agent-teams-vs-arch2
Anthropic agent-teams vs AI Uni Architecture 2 comparison
S74-3 · researcher · track: cross-track
Landscape brief: hermes-agent-ecosystem-evaluation
Hermes agent ecosystem evaluation
S74-4 · researcher · track: prototype
Landscape brief: kb-ia-visualization-patterns
KB IA visualization patterns research
S74-5 · researcher · track: cross-track
Landscape brief: lane1-pattern-catalog-and-pipeline
Pattern catalog + 4-stage pipeline + two-agent split
S74-6 · researcher · track: cross-track
Landscape brief: lane5-phase-b-productization-and-q2-amendment
Phase 1.5 + Phase 4b NEW + 6 D1-D6 decisions
Sibling landscape briefs (3)
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.
LangSmith assessment brief (SO #31)
Recommendation: deterministic CI gates remain bash-fail-closed; LLM-as-judge evaluators are post-hoc observability only.
Opus 4.7 urgent assessment
Capabilities · pricing · 1M context · cache TTL implications. Standing P0 since S56.
Future artifacts · pending (11)
F-2 · track: cross-track
QA Uplift L-5 corpus
Anticipated location: docs/agent-knowledge/qa/expertise/
F-3 · track: cross-track
Validator Uplift L-6 corpus
Anticipated location: docs/agent-knowledge/validator/expertise/
F-4 · track: cross-track
Code-PM Uplift L-7 corpus
Anticipated location: docs/agent-knowledge/code-pm/expertise/
F-5 · track: cross-track
PO Uplift L-8 corpus
Anticipated location: docs/agent-knowledge/po/expertise/
F-12 · track: cross-track
Auditor Uplift L-12 corpus
Anticipated location: docs/agent-knowledge/auditor/expertise/
F-13 · track: cross-track
Administrator Uplift L-13 corpus
Anticipated location: docs/agent-knowledge/administrator/expertise/
F-15 · track: cross-track
LENS-SUITE-REGISTRY 8-class expansion (M-4)
Anticipated location: docs/product-requirements/LENS-SUITE-REGISTRY.md (amended)
F-16 · track: cross-track
M-5 quarterly refresh cadence design
Anticipated location: docs/agent-knowledge/quarterly-refresh-cadence.md (TBD)
F-17 · track: prototype
KB v4 Designer HANDOFF_PACKET
Anticipated location: docs/product-requirements/KB-V4-DESIGNER-HANDOFF-PACKET-S75plus.md
F-18 · track: prototype
KB v4 Designer deliverables
Anticipated location: KB v4 prototype worktree (TBD)
F-19 · track: prototype
KB v4 implementation + deploy
Anticipated location: KB v4 build worktree (TBD)