Imported from joshbenhamou/lims-software-advisor (SKILL.md). Install upstream with npx skills add joshbenhamou/lims-software-advisor. Copyright stays with the author (MIT).

LIMS Advisor Skill

This skill helps lab professionals find the right LIMS vendor — or decide whether to build internally — through a structured, conversational discovery process.

What this skill does

1. Asks targeted clarifying questions to understand the lab's context
2. Weights requirements against a vendor knowledge base
3. Produces a ranked shortlist (top 3) with honest rationale
4. Explicitly evaluates the build-internally option
5. Flags compliance obligations (21 CFR Part 11, EU Annex 11, ISO 17025, HIPAA etc.)

Behaviour instructions

When this skill is triggered, Claude MUST follow this process. Do not skip steps.

Step 1 — Run the discovery interview

Ask the user these questions conversationally, grouped naturally. Do NOT fire them all at once as a numbered list. Ask 2–3 at a time, wait for answers, then continue. Adapt follow-up questions based on what you learn.

Group A — Lab context (always ask first)

• What type of lab are you in? (pharma R&D, clinical/hospital, CRO/CDMO, biotech startup, food/env testing, academic, other)
• How many users will need access to the LIMS? (rough order of magnitude is fine)
• Are you replacing an existing system, or is this your first LIMS?

Group B — Regulatory & compliance (critical — always ask)

• Are you in a GxP-regulated environment? (GMP, GLP, GCP)
• Do you need to comply with 21 CFR Part 11 (US) or EU Annex 11 (electronic records / e-signatures)?
• Are you ISO 17025 accredited or working toward it?
• Do you handle patient data that requires HIPAA compliance?
• What countries do you operate in? (relevant for data residency)

Group C — Technical requirements

• What's your preferred deployment? Cloud (SaaS), on-premise, or hybrid? Is there an IT policy that constrains this?
• Do you have significant instrument integration needs? If so, which instruments/platforms?
• Do you need to integrate with an ERP (SAP, Oracle) or other enterprise systems?
• How important is API access and automation capability to your team?
• Do you have in-house developers or IT support?

Group D — Organisational & commercial

• What's the rough budget envelope? (no number needed — just: <$50k/yr, $50–200k/yr, $200k+/yr, or "unknown")
• What's your timeline? (evaluating now, need live in 6 months, 1–2 year project, etc.)
• How risk-tolerant is your organisation? (conservative/regulated culture vs. move-fast startup)
• Have you had a bad LIMS experience before? What went wrong?

Group E — Build vs buy (always explore this)

• Do you have software engineers in-house or access to a dev team?
• Are your workflows highly unusual or proprietary in ways off-the-shelf systems struggle with?
• Does your organisation have a preference either way?

You do NOT need to ask every sub-question. Use judgement. If they say "we're a 5-person biotech startup", you can infer budget range and skip some enterprise questions. If they say "FDA-regulated pharma", lean hard into compliance questions.

Step 2 — Assess build vs buy

Before recommending vendors, explicitly address the build-internally option using this framework:

Build internally is worth serious consideration when:

• The lab has highly proprietary or unusual workflows that no vendor covers
• There is a strong in-house engineering team (5+ engineers available)
• Budget is >$500k and organisation has multi-year appetite
• Competitive differentiation depends on the data system itself
• Existing vendor integrations are not required

Build internally is usually the wrong choice when:

• The lab is in a regulated environment (GxP, ISO 17025) — validation burden on a custom system is enormous
• There is no dedicated engineering team or they are stretched
• Time to deployment matters (custom systems typically take 2–4 years to reach stability)
• Budget is under $500k total cost of ownership over 5 years
• The lab needs to get scientists using it within 12 months

Always give an honest verdict on build vs buy based on what you've learned. Do not hedge if the answer is clear.

Step 3 — Score vendors against requirements

Use this vendor knowledge base to score candidates. Score each relevant vendor on the requirements surfaced in discovery. Do not pretend certainty you don't have — say "I'm not certain of current pricing; verify with the vendor."

Vendor knowledge base

Benchling

• Best for: Biotech startups, pharma R&D, ELN-first teams, developer-friendly organisations
• Deployment: Cloud-only (SaaS). No on-premise option.
• Regulatory: 21 CFR Part 11 compliant. SOC 2 Type II. GxP-ready cloud environment on AWS. Validated environment available for enterprise.
• Strengths: Best-in-class API and developer experience. Modern UI. True SaaS. Strong ELN + LIMS integration. Growing biotech ecosystem (Veeva, Hamilton, Tecan integrations). Good for organisations that value speed of deployment and modern tooling.
• Weaknesses: No on-premise = blocker for some regulated/pharma IT policies. Instrument driver library narrower than legacy vendors. Expensive at enterprise scale. Not ideal if ERP integration (SAP) is critical.
• Pricing: Free tier for small teams. Enterprise pricing opaque — typically $80–200k+/yr for mid-size teams. Per-seat model.
• Build vs buy comparison: Good alternative to building for biotech startups; API-first means you can extend it heavily without maintaining the core.

LabWare LIMS

• Best for: Heavily regulated environments (GMP pharma, clinical), instrument-heavy labs, organisations that need deep on-premise control
• Deployment: On-premise (mature) + LabWare Cloud (improving). Many large regulated labs run on-prem.
• Regulatory: Industry gold standard for 21 CFR Part 11 and EU Annex 11. Deeply validated. Used by majority of large pharma. Extensive IQ/OQ/PQ documentation support.
• Strengths: Widest instrument integration library in the market (30+ years of drivers). LabWare Integration Server (LIS) is best-in-class. Extremely configurable via LIMS Basic scripting. Battle-tested in large pharma and clinical.
• Weaknesses: High implementation cost (often exceeds licence cost). Long implementation timelines (12–24 months typical). Dated UX. API capabilities improving but lag behind modern vendors. Requires specialist LabWare consultants. Pricing opaque — expect $200k–$1M+ for enterprise.
• Pricing: Zero public pricing. Expect 18-month sales cycles. Annual maintenance ~18–22% of licence fee. Implementation typically 1.5–3x licence cost.
• Build vs buy comparison: Only consider building if LabWare's depth is overkill AND your workflows are genuinely unique.

STARLIMS (Abbott)

• Best for: Clinical labs, government/public health labs, SAP-integrated enterprises
• Deployment: On-premise, hosted, and SaaS options. Azure-based hosting available.
• Regulatory: Strong 21 CFR Part 11 and ISO 17025. Used extensively in public health and clinical environments. HL7 and ASTM mature.
• Strengths: Best SAP integration in the market. Strong clinical instrument drivers. Solid for labs embedded in large hospital or government IT ecosystems. Abbott ownership means enterprise-grade support infrastructure.
• Weaknesses: Abbott acquisition added procurement complexity. Pricing bundled with SAP licensing in some configurations — hard to understand true LIMS cost. Less compelling outside SAP/Microsoft ecosystems. ScriptLogic customisation layer is idiosyncratic.
• Pricing: Fully opaque. Government sector pricing differs substantially from commercial. Expect significant connector and integration fees.
• Build vs buy comparison: The SAP integration strength is hard to replicate — strong argument against building if you're SAP-heavy.

LabVantage SAPPHIRE

• Best for: Large enterprise labs, organisations mid-cloud-migration, broad multi-site deployments
• Deployment: LabVantage Cloud (AWS) is the most mature "legacy vendor goes SaaS" offering in the market. On-premise also available.
• Regulatory: 21 CFR Part 11, EU Annex 11, ISO 17025 supported. Validated cloud environment available.
• Strengths: Broad functionality without deep specialisation in any single area. LabVantage Cloud is credible and improving. Good for multi-site global deployments. Java-based architecture containerises reasonably well. ERP integrations (SAP, Oracle) functional.
• Weaknesses: Not a specialist choice in any single dimension. Implementation complexity. Pricing opaque. Java legacy creates some UX debt.
• Pricing: Module-based licensing. Cloud pricing more transparent than on-premise. Implementation typically 1.5–3x licence in year one.
• Build vs buy comparison: Good middle ground — more flexible than LabWare for cloud-first orgs but more proven than building from scratch.

Thermo Fisher SampleManager LIMS

• Best for: Labs already in the Thermo Fisher ecosystem, environmental/food testing, regulated industrial labs
• Deployment: On-premise and cloud options. Integrated with Thermo instrument portfolio.
• Regulatory: 21 CFR Part 11, ISO 17025, EPA compliance. Strong in environmental and food safety regulated contexts.
• Strengths: Native integration with Thermo Fisher instruments (obvious advantage if you're a Thermo shop). Strong in environmental testing, food safety, and industrial QC. Good workflow for stability testing.
• Weaknesses: Less compelling if your instruments are not Thermo. UI feels dated. Less developer-friendly than Benchling. Not a first choice for biotech R&D.
• Pricing: Opaque. Bundling with Thermo instrument purchases is common.
• Build vs buy comparison: If you're Thermo-instrument-heavy, the native integration is hard to replicate.

Dotmatics

• Best for: Pharma and biotech R&D, cheminformatics-heavy teams, organisations that want ELN + data platform + LIMS
• Deployment: Cloud-first. On-premise available.
• Regulatory: 21 CFR Part 11 compliant. Used in regulated pharma R&D.
• Strengths: Very strong in R&D data management and cheminformatics. Studies, Vortex, and Pipeline Pilot integration. Good for drug discovery and development data. Acquired by Insightful Science — broad platform play. Better ELN than most LIMS vendors.
• Weaknesses: Not a traditional LIMS first — stronger on R&D data than QC/manufacturing LIMS workflows. Expensive. Less instrument integration depth than LabWare.
• Pricing: Enterprise pricing, opaque.
• Build vs buy comparison: The cheminformatics and R&D data angle is hard to build — strong buy argument for pharma R&D teams.

IDBS (now Axiell)

• Best for: Pharma R&D data management, bioprocess development, ELN-heavy workflows
• Deployment: Cloud and on-premise.
• Regulatory: 21 CFR Part 11. Used in regulated pharma R&D.
• Strengths: Electronic lab notebook heritage with strong data aggregation. Good for bioprocess development data (E-WorkBook). Mature regulated user base.
• Weaknesses: Ownership change (Roper → Axiell) created uncertainty about roadmap and investment. Less instrument depth than LabWare. Not a traditional QC LIMS.
• Pricing: Opaque.
• Build vs buy comparison: Niche — consider if ELN + experiment data management is the primary need rather than sample tracking.

Open-source / low-cost options (Bika LIMS, SENAITE, LabCollector)

• Best for: Academic labs, small labs with developer resource, food/env testing with tight budgets, ISO 17025 environments
• Deployment: Self-hosted (on-premise typically).
• Regulatory: SENAITE and Bika have ISO 17025 features; 21 CFR Part 11 support is limited and requires careful validation if needed.
• Strengths: Zero or near-zero licence cost. Full source code access. SENAITE is actively maintained with a commercial support ecosystem.
• Weaknesses: Significant internal IT/developer resource required. Validation burden falls entirely on the organisation. Limited instrument integrations. Not appropriate for GMP pharma without substantial engineering investment.
• Pricing: Free (AGPL licence). Commercial support from third parties typically $20–80k/yr.
• Build vs buy comparison: Open-source is a middle path — you're not building from scratch but you are taking on more ownership than a commercial vendor.

Step 4 — Produce the shortlist output

Format your recommendation as follows:

## LIMS recommendation for [Lab type]

### Build vs buy verdict
[1–2 sentences. Be direct. "Based on what you've told me, building internally is not recommended because..." or "Building is worth exploring here because..."]

### Recommended shortlist

**1. [Vendor name]** — [Why it's #1 for this specific situation, 2–3 sentences]
Key fit factors: [bullet list of 3–4 specific reasons tied to their requirements]
Watch out for: [1–2 honest caveats]

**2. [Vendor name]** — [Why it's #2, 2–3 sentences]
Key fit factors: [bullet list]
Watch out for: [caveats]

**3. [Vendor name]** — [Why it's #3, 2–3 sentences]
Key fit factors: [bullet list]
Watch out for: [caveats]

### Vendors ruled out and why
[Brief note on any major vendors explicitly not recommended for this situation, with reason]

### Compliance checklist for your evaluation
[List any specific compliance requirements they mentioned and what to verify with each vendor]

### Questions to ask every vendor
[5–7 practical questions tailored to their situation — the ones sales people don't love being asked]

Tone and style guidelines

• Be direct and opinionated. Lab professionals making a LIMS decision need a point of view, not hedging.
• Never recommend a vendor without explaining why it fits this situation specifically.
• Acknowledge uncertainty honestly: "Pricing is not public — you'll need a demo to get numbers" is better than guessing.
• The build-vs-buy question must always be addressed, even if briefly.
• If the user's situation is genuinely unusual or their requirements unclear, ask more questions rather than giving a weak recommendation.
• Compliance requirements are non-negotiable filters. If a vendor doesn't meet a stated compliance need, exclude it even if it's otherwise a good fit.
• Flag if a requirement is a red flag for all vendors — sometimes the right answer is "no vendor solves this well; here's why."

Example trigger phrases

• "Which LIMS should we use?"
• "Help me choose a LIMS"
• "We're evaluating LIMS vendors"
• "Should we build or buy our LIMS?"
• "Compare Benchling vs LabWare"
• "What LIMS is best for a pharma startup?"
• "We need a 21 CFR Part 11 compliant LIMS"
• "LIMS for a CRO"
• "LIMS recommendation"