Plan clinical trials and research budgets

Design clinical studies and manage research program budgets

Research teams spanning clinical, finance, market, and product domains when comprehensive research-ops infrastructure is needed across domains.

• · Research idea (clinical, finance, market, or product-focused)
• · Career stage, preliminary data, organizational environment
• · Submission posture (grant/internal funding/partner)
• · Domain-specific context (known institute targets, budget constraints)

• · Study design artifacts (protocol, endpoints, sample-size/power, phase-gating)
• · Research finance outputs (R&D program budgeting, F&A modeling, capitalize-vs-expense routing)
• · Market research outputs (TAM/SAM/SOM, survey design, segmentation)
• · Product research outputs (interview design, saturation, insight repository)

• · 24 stdlib-only Python tools (12 analysis + 12 onboarding/customization/autoresearch)
• · onboard.py, config_loader.py (customization per sub-skill)
• · ar_evaluator.py (bridges to engineering/autoresearch-agent)

• · Each sub-skill (clinical, finance, market, product) has first-run onboarding
• · Customization loader reads per-skill config files (project > global > defaults precedence)
• · Autoresearch evaluator is opt-in and isolated (ground-truth tool)

• · Sub-skills require user expertise per domain (e.g., clinical design needs statistical knowledge)
• · Config precedence (project > global > defaults) can mask missing configs (verify on first run)
• · Autoresearch evaluator must be manually trusted before enabling (no auto-trust)

• · 5 sub-skills verified by implementation plan (research-ops-expansion-plan.md)
• · 24 stdlib-only tools are deterministic (no external APIs)
• · Forcing-question library per sub-skill (Matt Pocock grill-with-docs discipline)
• · Config loader + autoresearch evaluator are explicit integration points
• · Distinct from ra-qm-team (regulatory), finance (corporate), research/grants (NIH discovery)