agentleFS
aws-samplesaws-samples/amazon-bedrock-agents-healthcare-lifesciences/skills/clinical-protocol-drafting/SKILL.md

clinical-protocol-drafting

Skill274 starsChanged 6 months agoMIT-0View on GitHub

Use when drafting clinical trial protocol sections (objectives, background, study design) grounded in ICH guidelines (E6, E8, E9) and FDA regulations (21 CFR Part 312). Generates IRB-ready protocol documents from grant documents or study synopses.

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 5 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Ensure protocol language complies with ICH E6(R2), E8(R1), E9

    line 12
  2. Distinctive:

    Check regulatory requirements from 21 CFR Part 312 for IND protocols

    line 13
  3. Very specific:

    Use retrieve with query "protocol objectives requirements ICH E6(R2) Section 6" to retrieve guidance on how protocol objectives should be …

    line 34
  4. Very specific:

    Use get_cfr_section with part "312" and section "23" to retrieve 21 CFR 312.23 requirements on nonclinical and clinical background …

    line 48
  5. Very specific:

    Use retrieve with query "protocol background section requirements ICH E6(R2) E8(R1)" to retrieve guidance.

    line 49
  6. Very specific:

    Use retrieve with query "general study design considerations randomization ICH E8(R1)" for design guidance.

    line 66
  7. Very specific:

    Use retrieve with query "statistical design principles sample size ICH E9" for statistical principles.

    line 67
  8. Very specific:

    Use get_cfr_section with part "312" and section "23" for 21 CFR 312.23(a)(6) requirements on protocol design elements.

    line 68
  9. Distinctive:

    Do not alter scientific content.

    line 90
  10. Distinctive:

    Use ICH-compliant terminology throughout (e.g., "investigational medicinal product" not "drug")

    line 95
  11. Distinctive:

    Never fabricate clinical data — use only what is provided in the source document

    line 99

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Clinical Trial Protocol Drafting

When to use this skill

  • Draft protocol sections from a grant document or study synopsis
  • Generate objectives, background/rationale, or study design sections
  • Ensure protocol language complies with ICH E6(R2), E8(R1), E9
  • Check regulatory requirements from 21 CFR Part 312 for IND protocols
  • Assemble multiple sections into a cohesive protocol document

MCP Dependencies

MCP ServerToolPurpose
fda-ecfrget_cfr_sectionRetrieve 21 CFR regulatory text
fda-ecfrsearch_cfrSearch across CFR titles
awslabs.bedrock-kb-retrieval-mcp-serverretrieveQuery ICH E6/E8/E9 content via Bedrock Knowledge Base

The fda-ecfr server is in mcp-servers/agentcore-gateway/fda-ecfr/ — deploy it first. For ICH guidelines, set up a Bedrock Knowledge Base per references/ich-knowledge-base-setup.md and use the existing awslabs bedrock-kb-retrieval-mcp-server.

Workflow: Protocol Drafting from Grant Document

Step 1: Objectives Section

Generate the draft Objectives section of a clinical trial protocol.

  1. Extract from the grant: specific aims, primary endpoint, secondary endpoints.
  2. Use retrieve with query "protocol objectives requirements ICH E6(R2) Section 6" to retrieve guidance on how protocol objectives should be stated.

Output structure:

  • Primary Objective (linked to the primary endpoint)
  • Secondary Objectives (linked to each secondary endpoint)
  • Exploratory Objectives (if any are implied by the grant aims)

Each objective: clear, measurable, single-sentence statement following ICH conventions. Map each objective to its corresponding endpoint.

Step 2: Background & Rationale Section

Generate the draft Background and Rationale section.

  1. Extract from the grant: disease context, preliminary data (preclinical and Phase 1 results), unmet medical need, scientific rationale for the investigational agent.
  2. Use get_cfr_section with part "312" and section "23" to retrieve 21 CFR 312.23 requirements on nonclinical and clinical background information for an IND protocol.
  3. Use retrieve with query "protocol background section requirements ICH E6(R2) E8(R1)" to retrieve guidance.

Output structure:

  • Disease Overview (epidemiology, molecular subtype prevalence, current standard of care)
  • Unmet Medical Need (limitations of existing therapies)
  • Investigational Agent Summary (mechanism of action, selectivity profile)
  • Relevant Nonclinical Findings (IC50 data, selectivity over wild-type)
  • Clinical Experience to Date (Phase 1 dose-escalation results, RP2D, preliminary efficacy and safety signals)
  • Study Rationale (why this agent, this population, this design)

Tone: scientific, appropriate for IRB submission. Cite grant preliminary data where applicable.

Step 3: Study Design Section

Generate the draft Study Design section.

  1. Extract from the grant: study design type, randomization scheme, treatment arms, dosing, sample size, study duration, number of sites, target population.
  2. Use retrieve with query "general study design considerations randomization ICH E8(R1)" for design guidance.
  3. Use retrieve with query "statistical design principles sample size ICH E9" for statistical principles.
  4. Use get_cfr_section with part "312" and section "23" for 21 CFR 312.23(a)(6) requirements on protocol design elements.

Output structure:

  • Overall Design (phase, randomization ratio, open-label justification, multicenter)
  • Treatment Arms (experimental arm with dose/schedule, control arm with regimen options)
  • Study Schema (text-based visual flow: screening → randomization → treatment → follow-up)
  • Randomization and Stratification (method, stratification factors if applicable)
  • Sample Size Justification (statistical basis, power, expected effect size)
  • Study Duration (enrollment period, treatment duration, follow-up period)

Step 4: Assembly

Assemble all drafted sections into a single cohesive protocol document.

Formatting:

  • Protocol title page (study title, PI, funding source, protocol version/date)
  • Sequential section numbering
  • Consistent formatting, tense, and terminology
  • Table of Contents
  • Cross-reference harmonization (objectives referenced in study design)
  • Flag inconsistencies as reviewer comments

Do not alter scientific content. Only harmonize language and ensure smooth transitions.

Conventions

  • Lead with the most clinically relevant information in each section
  • Use ICH-compliant terminology throughout (e.g., "investigational medicinal product" not "drug")
  • Distinguish primary from secondary objectives clearly
  • Include regulatory citations inline (e.g., "per 21 CFR 312.23(a)(6)")
  • For VUS or uncertain data: note limitations and recommend additional review
  • Never fabricate clinical data — use only what is provided in the source document
  • Flag sections where the grant provides insufficient detail for protocol-level specificity

Regulatory References

  • ICH E6(R2): Good Clinical Practice, Integrated Addendum
  • ICH E8(R1): General Considerations for Clinical Studies
  • ICH E9: Statistical Principles for Clinical Trials
  • 21 CFR Part 312: Investigational New Drug Application
  • 21 CFR Part 50: Protection of Human Subjects
  • 21 CFR Part 56: Institutional Review Boards

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.