hh-health-ai/healthcare-equity/device-diligence
Claude Code and agent skill by hh-health-ai/healthcare-equity
- community
Checking that the SKILL.md file is still in the repository.
This skill should be used when the user asks about "510(k) predicate risk", "De Novo vs PMA pathway", "FDA meeting history for [asset]", "TPP assessment", "CMC/manufacturing scalability", "freedom to operate / IP risk", "LDT rule impact", or "companion diagnostic linkage" — the regulatory and technical diligence layer for drugs, devices, and diagnostics.
Install
npx vskill@latest install hh-health-ai/healthcare-equity/device-diligenceSKILL.md
Device & Regulatory Diligence
Deep regulatory and technical diligence: product profile, manufacturing, IP, FDA pathway and history, diagnostics regulation, and therapy–test coupling. Eight workflows, each a prompt in ${CLAUDE_PLUGIN_ROOT}/skills/device-diligence/references/prompts.md — pick by question:
| Question | Prompt | Primary sources | |
|---|---|---|---|
| Is the product profile differentiated or me-too? | SUB-BIO-04 (TPP interrogation) | Corporate deck, protocol, comparator labels | |
| Can they make it at scale, at margin? | SUB-BIO-05 (CMC/COGS scalability) | Filings, manufacturing disclosures, CDMO contracts | |
| Is the IP position durable? | SUB-BIO-06 (FTO analysis) | Patent databases, Orange/Purple Book, Para IV notices | |
| What does the FDA interaction history signal? | SUB-BIO-07 (meeting history risk) | SEC filings, Drugs@FDA, correspondence disclosures | |
| Is the 510(k) predicate sound? | SUB-MED-03 (predicate validity) | FDA 510(k) database, classification regs | |
| De Novo or PMA — and what does the pathway cost? | SUB-MED-04 (pathway risk) | FDA classification, recent comparables | |
| What does the LDT rule do to this diagnostic? | SUB-TLS-03 (LDT oversight impact) | FDA LDT final rule, company disclosures, litigation status | |
| Does a companion diagnostic gate the therapy? | SUB-TLS-05 (CDx linkage) | Drugs@FDA labels, CDx approvals, guidelines | |
| Does a delivery-format change (IV→SC, autoinjector, PFS) carry the thesis? | Combination-product module (below) | FDA combination-product precedents, ISO 11608, DMEPA/URRA records |
Execution rules
- Choose the workflow(s) from the table, load the prompt, and gather primary sources first: FDA databases via web research (Drugs@FDA, 510(k)/PMA/De Novo databases, warning letters), the Clinical Trials connector for supporting studies, PubMed for published validation data.
- Ground pathway and development-stage judgments in
${CLAUDE_PLUGIN_ROOT}/references/fda-pathways-fto.mdand${CLAUDE_PLUGIN_ROOT}/references/device-development-checkpoints.md(includes the human-factors gate). For drug CMC, translate per the manufacturing row of references/evidence-translation-clinical.md: CMC progress moves launch probability, timing, yield, capacity, and capital — not patient demand or share. - For legal-adjacent outputs (FTO), state the analysis is an investment-diligence screen, not a legal opinion, and note that definitive FTO requires patent counsel.
- End each workflow with the EVIDENCE BRIEF block (layer: regulatory; competitive for TPP/CDx where share-relevant). Route label/coverage consequences to adcom-label and cms-reimbursement; route pathway timelines to catalyst-calendar; route COGS/margin deltas to healthcare-equity model-valuation.
Combination products & delivery-format changes
For IV→SC conversions, autoinjector/PFS launches, self-administration claims, and device changes on marketed drugs, work from ${CLAUDE_PLUGIN_ROOT}/references/combination-products.md: place the change on the four-domain evidence ladder (HF/URRA · PK bridging 80–125% · clinical · CMC/EDDO, with the Humira-s381 → Repatha precedent range), check device qualification (ISO 11608, aged/shipped verification, PPQ at filing) and the autoinjector deficiency base rates (~10/12 recent BLAs flagged on essential-performance traceability), note that shelf-life/device-change supplements are PAS-clocked dated events, and translate per the "easier administration" evidence row — a format conversion moves initiation/site-of-care/persistence, never efficacy or TAM, and is not a new patient (case C05 rule). Post-launch, hand MAUDE use-error trajectories to the quality-signals skill.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingmodules/clinical-catalysts/skills/device-diligence/SKILL.md
Popularity
- Repository stars
- 101
- Repository forks
- 8
- 7-day trend score
- 0
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/hh-health-ai/healthcare-equity/device-diligence)