[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M73FDWT13QHYTHGXQWD5KJPB":3,"skill-same:01M73FDWT13QHYTHGXQWD5KJPB":97},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":16,"sourceHtmlUrl":17,"categories":18,"keywords":19,"publishedAt":20,"linkOnly":21,"markdownContent":22,"markdownBody":23,"frontMatter":24,"refName":25,"commitSha":26,"skillDirectory":5,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":30,"ownerDescription":6,"updatedAt":31,"install":32,"resources":45},"01M73FDWT13QHYTHGXQWD5KJPB","clinical-trial-protocol-skill",null,"Generate clinical trial protocols for medical devices or drugs. This skill should be used when users say \"Create a clinical trial protocol\", \"Generate protocol for [device\u002Fdrug]\", \"Help me design a clinical study\", \"Research similar trials for [intervention]\", or when developing FDA submission documentation for investigational products.","anthropics\u002Flife-sciences","anthropics","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F76263028?v=4","01M697XGDPXJAVQ8391V2BCYTE","Anthropic","01M718E634A37G82YPYC04ZY81",651,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: clinical-trial-protocol-skill\ndescription: Generate clinical trial protocols for medical devices or drugs. This skill should be used when users say \"Create a clinical trial protocol\", \"Generate protocol for [device\u002Fdrug]\", \"Help me design a clinical study\", \"Research similar trials for [intervention]\", or when developing FDA submission documentation for investigational products.\n---\n\n# Clinical Trial Protocol Skill\n\n## ⚠️ EXECUTION CONTROL - READ THIS FIRST\n\n**CRITICAL: This orchestrator follows a SIMPLE START approach:**\n\n1. **Display the welcome message FIRST** (shown in \"Startup: Welcome and Confirmation\" section below)\n2. **Ask user to confirm they're ready to proceed** - Wait for confirmation (yes\u002Fno)\n3. **Jump directly into Full Workflow Logic** - Automatically run subskills sequentially\n4. **Do NOT pre-read subskill files** - Subskills are loaded on-demand only when their step executes\n\n**Why this matters:**\n- Pre-reading all subskills wastes context and memory\n- Subskills should only load when actually needed during execution\n- Workflow automatically handles resuming from existing waypoints\n\n## Overview\n\nThis skill generates clinical trial protocols for **medical devices or drugs** using a **modular, waypoint-based architecture** \n\n## What This Skill Does\n\nStarting with an intervention idea (device or drug), this orchestrated workflow offers two modes:\n\n**🔬 Research Only Mode (Steps 0-1):**\n0. **Initialize Intervention** - Collect device or drug information\n1. **Research Similar Protocols** - Find similar trials, FDA guidance, and published protocols\n   - **Deliverable:** Comprehensive research summary as formatted .md artifact\n\n**📄 Full Protocol Mode (Steps 0-5):**\n0. **Initialize Intervention** - Collect device or drug information\n1. **Research Similar Protocols** - Find similar trials, FDA guidance, and published protocols\n2. **Protocol Foundation** - Generate protocol sections 1-6 (foundation, design, population)\n3. **Protocol Intervention** - Generate protocol sections 7-8 (intervention details)\n4. **Protocol Operations** - Generate protocol sections 9-12 (assessments, statistics, operations)\n5. **Generate Protocol** - Create professional file ready for stakeholder review\n\n## Architecture\n\n### Waypoint-Based Design\n\nAll analysis data is stored in `waypoints\u002F` directory as JSON\u002Fmarkdown files:\n\n```\nwaypoints\u002F\n├── intervention_metadata.json           # Intervention info, status, initial context\n├── 01_clinical_research_summary.json   # Similar trials, FDA guidance, recommendations\n├── 02_protocol_foundation.md            # Protocol sections 1-6 (Step 2)\n├── 03_protocol_intervention.md          # Protocol sections 7-8 (Step 3)\n├── 04_protocol_operations.md            # Protocol sections 9-12 (Step 4)\n├── 02_protocol_draft.md                 # Complete protocol (concatenated in Step 4)\n├── 02_protocol_metadata.json            # Protocol metadata\n└── 02_sample_size_calculation.json      # Statistical sample size calculation\n```\n\n**Rich Initial Context Support:**\nUsers can provide substantial documentation, technical specifications, or research data when initializing the intervention (Step 0). This is preserved in `intervention_metadata.json` under the `initial_context` field. Later steps reference this context for more informed protocol development.\n\n### Modular Subskill Steps\n\nEach step is an independent skill in `references\u002F` directory:\n\n```\nreferences\u002F\n├── 00-initialize-intervention.md    # Collect device or drug information\n├── 01-research-protocols.md         # Clinical trials research and FDA guidance\n├── 02-protocol-foundation.md        # Protocol sections 1-6 (foundation, design, population)\n├── 03-protocol-intervention.md      # Protocol sections 7-8 (intervention details)\n├── 04-protocol-operations.md        # Protocol sections 9-12 (assessments, statistics, operations)\n└── 05-generate-document.md          # NIH Protocol generation\n```\n\n### Utility Scripts\n\n```\nscripts\u002F\n├── sample_size_calculator.py        # Statistical power analysis\n└── test_sample_size_calculator.py   # Unit tests for the calculator\n```\n\n## Prerequisites\n\n### 1. clinical trials MCP Server (Required)\n\n**Installation:**\n- Install via drag-and-drop `.mcpb` file into Claude Desktop\n- Or configure manually in Claude Desktop settings\n\n**Available Tools:**\n`search_clinical_trials` - Search by:\n\ncondition - Disease or condition (e.g., \"pancreatic cancer\")\nintervention - Drug, device, or treatment (e.g., \"pembrolizumab\", \"CAR-T\")\nsponsor - Sponsor or collaborator name (e.g., \"Pfizer\", \"NIH\")\nlocation - City, state, or country (e.g., \"California\", \"Boston\")\nstatus - \"recruiting\" (default), \"active\", \"completed\", \"all\"\nphase - Trial phase: \"1\", \"2\", \"3\", \"4\", \"early_phase1\"\nmax_results - Default 25, max 100\n\n\n`get_trial_details` - Get comprehensive details for a specific trial using its nct_id (e.g., \"NCT04267848\"). Returns eligibility criteria, outcomes, study design, and contact information.\n\n**Verification:** Step 1 will automatically test MCP connectivity at startup.\n\n### 2. FDA Database Access (Built-in)\n\n**Purpose:** FDA regulatory pathway research via explicit database URLs\n\n**Sources:**\n- Step 1: FDA device\u002Fdrug databases (510(k), PMA, De Novo, Drugs@FDA, Orange Book, Purple Book)\n- All sources use direct FDA database URLs - no generic web searches\n\n### 3. Clinical Protocol Template\n\n**Template Files:** Any `.md` files in the `assets\u002F` directory\n\n**Purpose:** Reference template for protocol structure and content guidance. The system automatically detects available templates and uses them dynamically.\n\n### 4. Python Dependencies (Required for Step 2)\n\n**Installation:**\n```bash\npip install -r requirements.txt\n```\n\n**Dependencies:**\n- scipy >= 1.11.0 (statistical calculations)\n- numpy >= 1.24.0 (numerical operations)\n\n**Purpose:** Statistical sample size calculations for clinical protocols\n\n## How to Use\n\nSimply invoke the skill and select your desired mode:\n\n**🔬 Research Only Mode:**\n1. Select \"Research Only\" from the main menu\n2. Provide intervention information\n3. Receive comprehensive research summary as formatted .md artifact\n4. Option to continue with full protocol generation or exit\n\n**📄 Full Protocol Mode:**\n1. Select \"Full Protocol\" from the main menu\n2. Guide you through all steps sequentially (Steps 0-5)\n3. Pause after Step 4 to review the draft protocol\n4. Generate the final protocol document when ready\n\n**Resume Capability:** If interrupted, simply restart the skill and it will automatically resume from your last completed step.\n\n## Execution Flow\n\n### Startup: Welcome and Mode Selection\n\nWhen skill is invoked, display the following message:\n\n```\n🧬 CLINICAL TRIAL PROTOCOL\n\nWelcome! This skill generates clinical trial protocols for medical devices or drugs.\n\n[If waypoints\u002Fintervention_metadata.json exists:]\n✓ Found existing protocol in progress: [Intervention Name]\n  Type: [Device\u002FDrug]\n  Completed: [List of completed steps]\n  Next: [Next step to execute]\n\n📋 SELECT MODE:\n\n1. 🔬 Research Only - Run clinical research analysis (Steps 0-1)\n   • Collect intervention information\n   • Research similar clinical trials\n   • Find FDA guidance and regulatory pathways\n   • Generate comprehensive research summary as .md artifact\n\n2. 📄 Full Protocol - Generate complete clinical trial protocol (Steps 0-5)\n   • Everything in Research Only, plus:\n   • Generate all protocol sections\n   • Create professional protocol document\n\n3. ❌ Exit\n\nPlease select an option (1, 2, or 3):\n```\n\n**🛑 STOP and WAIT for user selection (1, 2, or 3)**\n\n- If **1 (Research Only)**: Set `execution_mode = \"research_only\"` and proceed to Research Only Workflow Logic\n- If **2 (Full Protocol)**: Set `execution_mode = \"full_protocol\"` and proceed to Full Workflow Logic\n- If **3 (Exit)**: Exit gracefully with \"No problem! Restart the skill anytime to continue.\"\n\n---\n\n### Research Only Workflow Logic\n\n**This workflow executes only Steps 0 and 1, then generates a formatted research summary artifact.**\n\n**Step 1: Check for Existing Waypoints**\n- If `waypoints\u002Fintervention_metadata.json` exists: Load metadata, check if steps 0 and 1 are already complete\n- If no metadata exists: Start from Step 0\n\n**Step 2: Execute Research Steps (0 and 1)**\n\nFor each step (0, 1):\n\n1. **Check completion status:** If step already completed in metadata, skip with \"✓ Step [X] already complete\"\n\n2. **Execute step:**\n   - Display \"▶ Executing Step [X]...\"\n   - Read and follow the corresponding subskill file instructions\n   - Wait for completion\n   - Display \"✓ Step [X] complete\"\n   - **Step execution method (ON-DEMAND LOADING):** When a step is ready to execute (NOT before), read the subskill markdown file and execute ALL instructions within it\n   - **Step-to-file mapping:**\n     - Step 0: `references\u002F00-initialize-intervention.md` (collect intervention info)\n     - Step 1: `references\u002F01-research-protocols.md` (clinical research and FDA guidance)\n\n3. **Handle errors:** If step fails, ask user to retry or exit. Save current state for resume capability.\n\n**Step 3: Generate Research Summary Artifact**\n\nAfter Step 1 completes successfully:\n\n1. **Read waypoint files:**\n   - `waypoints\u002Fintervention_metadata.json` (intervention details)\n   - `waypoints\u002F01_clinical_research_summary.json` (research findings)\n\n2. **Create formatted markdown summary:** Generate a comprehensive, well-formatted research summary as a markdown artifact with the following structure:\n\n```markdown\n# Clinical Research Summary: [Intervention Name]\n\n## Intervention Overview\n- **Type:** [Device\u002FDrug]\n- **Indication:** [Target condition\u002Fdisease]\n- **Description:** [Brief intervention description]\n- **Mechanism of Action:** [How it works]\n\n## Similar Clinical Trials\n[List top 5-10 similar trials with NCT ID, title, phase, status, key findings]\n\n## FDA Regulatory Pathway\n- **Recommended Pathway:** [510(k), PMA, De Novo, IND, NDA, BLA, etc.]\n- **Regulatory Basis:** [Rationale for pathway selection]\n- **Key Requirements:** [Major regulatory considerations]\n\n## FDA Guidance Documents\n[List relevant FDA guidance documents with links and key excerpts]\n\n## Study Design Recommendations\n- **Suggested Study Type:** [RCT, single-arm, etc.]\n- **Phase Recommendation:** [Phase 1, 2, 3, etc.]\n- **Primary Endpoint Suggestions:** [Based on similar trials]\n- **Sample Size Considerations:** [Preliminary thoughts]\n\n## Key Insights and Recommendations\n[Synthesized recommendations for protocol development]\n\n## Next Steps\n[If user wants to proceed with full protocol development]\n\n---\n*Generated by Clinical Trial Protocol Skill*\n*Date: [Current date]*\n```\n\n3. **Save artifact:** Write the formatted summary to `waypoints\u002Fresearch_summary.md`\n\n4. **Display completion message:**\n\n```\n✅ RESEARCH COMPLETE\n\nResearch Summary Generated: waypoints\u002Fresearch_summary.md\n\n📊 Key Findings:\n  • Similar Trials Found: [X trials]\n  • Recommended Pathway: [Pathway name]\n  • FDA Guidance Documents: [X documents identified]\n  • Study Design: [Recommended design]\n\n📄 The research summary has been saved as a formatted markdown artifact.\n\nWould you like to:\n1. Continue with full protocol generation (steps 2-5)\n2. Exit and review research summary\n\n```\n\n**Option 1 Logic (Continue to Full Protocol):**\n- Set `execution_mode = \"full_protocol\"`\n- Continue to Full Workflow Logic starting from Step 2 (since 0 and 1 are complete)\n\n**Option 2 Logic (Exit):**\n- Display: \"✓ Research summary saved. Restart the skill anytime to continue with protocol generation.\"\n- Exit orchestrator gracefully\n\n---\n\n### Full Workflow Logic\n\n**Step 1: Check for Existing Waypoints**\n- If `waypoints\u002Fintervention_metadata.json` exists: Load metadata, check `completed_steps` array, resume from next incomplete step\n- If no metadata exists: Start from Step 0\n\n**Step 2: Execute Steps in Order**\n\nFor each step (0, 1, 2, 3, 4, 5):\n\n1. **Check completion status:** If step already completed in metadata, skip with \"✓ Step [X] already complete\"\n\n2. **Execute step:** Display \"▶ Executing Step [X]...\", read and follow the corresponding subskill file instructions, wait for completion, display \"✓ Step [X] complete\"\n   - **Step execution method (ON-DEMAND LOADING):** When a step is ready to execute (NOT before), read the subskill markdown file and execute ALL instructions within it\n   - **IMPORTANT:** Do NOT read subskill files in advance. Only read them at the moment of execution.\n   - **Step-to-file mapping:**\n     - Step 0: `references\u002F00-initialize-intervention.md` (read when Step 0 executes)\n     - Step 1: `references\u002F01-research-protocols.md` (read when Step 1 executes)\n     - Step 2: `references\u002F02-protocol-foundation.md` (read when Step 2 executes - sections 1-6)\n     - Step 3: `references\u002F03-protocol-intervention.md` (read when Step 3 executes - sections 7-8)\n     - Step 4: `references\u002F04-protocol-operations.md` (read when Step 4 executes - sections 9-12)\n     - Step 5: `references\u002F05-concatenate-protocol.md` (read when Step 5 executes - final concatenation)\n\n3. **Handle errors:** If step fails, ask user to retry or exit. Save current state for resume capability.\n\n4. **Display progress:** \"Progress: [X\u002F6] steps complete\", show estimated remaining time\n\n5. **Step 4 Completion Pause:** After Step 4 completes, pause and display the Protocol Completion Menu (see below). Wait for user selection before proceeding.\n\n**Step 2.5: Protocol Completion Menu**\n\nAfter Step 4 completes successfully, display the EXACT menu below (do not improvise or create alternative options):\n\n```\n✅ PROTOCOL COMPLETE: Protocol Draft Generated\n\nProtocol Details:\n  • Study Design: [Design from metadata]\n  • Sample Size: [N subjects from metadata]\n  • Primary Endpoint: [Endpoint from metadata]\n  • Study Duration: [Duration from metadata]\n\nProtocol file: waypoints\u002F02_protocol_draft.md\nFile size: [Size in KB]\n\n📋 WHAT WOULD YOU LIKE TO DO NEXT?\n\n1. 📄 Review Protocol in Artifact - click on the .md file above\n\n2. 📄 Concatenate Final Protocol (Step 5)\n\n3. ⏸️  Exit and Review Later\n\n```\n\n**Option 1 Logic (Review in Artifact):**\nPause, let user open the section files, wait for further instruction\n\n**Option 2 Logic (Concatenate Protocol):**\n1. Execute Step 5 by reading and following `references\u002F05-concatenate-protocol.md`\n2. Step 5 will concatenate all section files into final protocol document\n3. Continue to Step 3 (Final Summary) after Step 5 completes\n\n**Option 3 Logic (Exit):**\n1. Display: \"✓ Protocol sections saved. You can resume with Step 5 anytime to concatenate.\"\n2. Exit orchestrator gracefully\n\n**Step 3: Final Summary**\n\nDisplay completion message with:\n- Intervention name, type (device\u002Fdrug), indication\n- Protocol details (design, sample size, endpoints, duration)\n- All completed steps list\n- Final deliverable: Complete protocol markdown file location (waypoints\u002Fprotocol_complete.md)\n- Waypoint files list for reference\n- Important disclaimers (FDA Pre-Sub, biostatistician review, IRB approval required)\n- Thank you message\n\n## Technical Details\n\n### Waypoint File Formats\n\n**JSON Waypoints** (Steps 0, 1):\n- Structured data for programmatic access\n- Small file sizes (1-15KB)\n- Easy to parse and reference\n\n**Markdown Waypoints** (Steps 2, 3, 4):\n- Step 2: `02_protocol_foundation.md` (Sections 1-6)\n- Step 3: `03_protocol_intervention.md` (Sections 7-8)\n- Step 4: `04_protocol_operations.md` (Sections 9-12)\n- Step 4: `02_protocol_draft.md` (concatenated complete protocol)\n- Human-readable protocol documents\n- Can be directly edited by users\n- Individual section files preserved for easier regeneration\n\n### Data Minimization Strategy\n\nEach step implements aggressive summarization:\n- **Keep:** Top-N results (5-10 max)\n- **Keep:** Key facts and IDs (NCT numbers, endpoint types)\n- **Keep:** Concise rationale (2-3 sentences)\n- **Discard:** Raw MCP query results (not needed after analysis)\n- **Discard:** Full FDA guidance text (only excerpts\u002Fcitations kept)\n- **Discard:** Lower-ranked search results\n\n### Step Independence\n\nEach subskill is designed to:\n- Read only from waypoint files (not conversation history)\n- Produce complete output in single execution\n- Not depend on conversation context from previous steps\n- Be runnable standalone\n\n## Error Handling\n\n### MCP Server Unavailable\n- Detected in: Step 1\n- Action: Display error with installation instructions\n- Allow user to retry after installing MCP server\n- No fallback available - MCP server is required for protocol research\n\n### Step Fails or Returns Error\n- Action: Display error message from subskill\n- Ask user: \"Retry step? (Yes\u002FNo)\"\n  - Yes: Re-run step\n  - No: Save current state, exit orchestrator\n\n### User Interruption\n- All progress saved in waypoint files\n- User can resume anytime by restarting the skill\n- Workflow automatically detects completed steps and resumes from next step\n- No data loss\n\n## Disclaimers\n\n⚠️ **IMPORTANT:** This protocol generation tool provides preliminary clinical study protocol based on NIH\u002FFDA guidelines and similar trials. It does NOT constitute:\n- Official FDA or IRB determination or approval\n- Medical, legal, or regulatory advice\n- Substitute for professional biostatistician review\n- Substitute for FDA Pre-Submission meeting\n- Guarantee of regulatory or clinical success\n\n**REQUIRED before proceeding with clinical study:**\n- Biostatistician review and sample size validation\n- FDA Pre-Submission meeting (Q-Submission for devices, Pre-IND for drugs)\n- IRB review and approval\n- Clinical expert and regulatory consultant engagement\n- Legal review of protocol and informed consent\n- Site investigator review and input\n- Sponsor completion of all [TBD] items in protocol\n\n**PROFESSIONAL CONSULTATION STRONGLY RECOMMENDED**\n\nClinical trial protocols are complex, high-stakes documents requiring expertise across multiple disciplines. Professional consultation with clinical trial experts, biostatisticians, and regulatory affairs specialists is essential before proceeding with clinical study planning.\n\n\n## Implementation Requirements\n\nWhen this skill is invoked:\n\n1. **Display the welcome message with mode selection** (shown in \"Startup: Welcome and Mode Selection\" section)\n\n2. **Wait for user mode selection** (1: Research Only, 2: Full Protocol, 3: Exit)\n\n3. **Execute based on selected mode:**\n   - **Research Only Mode (Option 1):**\n     - Execute Research Only Workflow Logic (Steps 0-1 only)\n     - Generate formatted research summary as .md artifact\n     - Offer option to continue with full protocol or exit\n   - **Full Protocol Mode (Option 2):**\n     - Execute Full Workflow Logic (Steps 0-5)\n     - Check for existing waypoints and resume from last completed step\n     - OR start from Step 0 if no waypoints exist\n     - Execute all steps sequentially until complete\n\n4. **For each step execution (LAZY LOADING - On-Demand Only):**\n   - **ONLY when a step is ready to execute**, read the corresponding subskill file\n   - Do NOT read subskill files in advance or \"to prepare\"\n   - Example: When Step 1 needs to run, THEN read `references\u002F01-research-protocols.md` and follow its instructions\n   - **For protocol development:** Execute Steps 2, 3, 4 sequentially in order\n   - Do NOT try to execute multiple steps in parallel - run sequentially\n   - Read each step's subskill file only when that specific step is about to execute\n\n5. **Research summary artifact generation (Research Only Mode):**\n   - After Step 1 completes, read waypoint files\n   - Generate comprehensive, well-formatted markdown summary\n   - Save to `waypoints\u002Fresearch_summary.md`\n   - Display completion message with key findings\n\n6. **Handle errors gracefully:**\n   - If a step fails, give user option to retry or exit\n   - If MCP server unavailable, explain how to install\n   - All progress is saved automatically in waypoints\n\n7. **Track progress:**\n   - Update `waypoints\u002Fintervention_metadata.json` after each step\n   - Show progress indicators to user (e.g., \"Progress: 3\u002F6 steps complete\" or \"Progress: 2\u002F2 research steps complete\")\n   - Provide clear feedback on what's happening\n\n8. **Final output:**\n   - **Research Only:** Display research summary location and offer to continue with full protocol\n   - **Full Protocol:** Congratulate user, display protocol location and next steps\n   - Remind user of disclaimers\n\n\n","# Clinical Trial Protocol Skill\n\n## ⚠️ EXECUTION CONTROL - READ THIS FIRST\n\n**CRITICAL: This orchestrator follows a SIMPLE START approach:**\n\n1. **Display the welcome message FIRST** (shown in \"Startup: Welcome and Confirmation\" section below)\n2. **Ask user to confirm they're ready to proceed** - Wait for confirmation (yes\u002Fno)\n3. **Jump directly into Full Workflow Logic** - Automatically run subskills sequentially\n4. **Do NOT pre-read subskill files** - Subskills are loaded on-demand only when their step executes\n\n**Why this matters:**\n- Pre-reading all subskills wastes context and memory\n- Subskills should only load when actually needed during execution\n- Workflow automatically handles resuming from existing waypoints\n\n## Overview\n\nThis skill generates clinical trial protocols for **medical devices or drugs** using a **modular, waypoint-based architecture** \n\n## What This Skill Does\n\nStarting with an intervention idea (device or drug), this orchestrated workflow offers two modes:\n\n**🔬 Research Only Mode (Steps 0-1):**\n0. **Initialize Intervention** - Collect device or drug information\n1. **Research Similar Protocols** - Find similar trials, FDA guidance, and published protocols\n   - **Deliverable:** Comprehensive research summary as formatted .md artifact\n\n**📄 Full Protocol Mode (Steps 0-5):**\n0. **Initialize Intervention** - Collect device or drug information\n1. **Research Similar Protocols** - Find similar trials, FDA guidance, and published protocols\n2. **Protocol Foundation** - Generate protocol sections 1-6 (foundation, design, population)\n3. **Protocol Intervention** - Generate protocol sections 7-8 (intervention details)\n4. **Protocol Operations** - Generate protocol sections 9-12 (assessments, statistics, operations)\n5. **Generate Protocol** - Create professional file ready for stakeholder review\n\n## Architecture\n\n### Waypoint-Based Design\n\nAll analysis data is stored in `waypoints\u002F` directory as JSON\u002Fmarkdown files:\n\n```\nwaypoints\u002F\n├── intervention_metadata.json           # Intervention info, status, initial context\n├── 01_clinical_research_summary.json   # Similar trials, FDA guidance, recommendations\n├── 02_protocol_foundation.md            # Protocol sections 1-6 (Step 2)\n├── 03_protocol_intervention.md          # Protocol sections 7-8 (Step 3)\n├── 04_protocol_operations.md            # Protocol sections 9-12 (Step 4)\n├── 02_protocol_draft.md                 # Complete protocol (concatenated in Step 4)\n├── 02_protocol_metadata.json            # Protocol metadata\n└── 02_sample_size_calculation.json      # Statistical sample size calculation\n```\n\n**Rich Initial Context Support:**\nUsers can provide substantial documentation, technical specifications, or research data when initializing the intervention (Step 0). This is preserved in `intervention_metadata.json` under the `initial_context` field. Later steps reference this context for more informed protocol development.\n\n### Modular Subskill Steps\n\nEach step is an independent skill in `references\u002F` directory:\n\n```\nreferences\u002F\n├── 00-initialize-intervention.md    # Collect device or drug information\n├── 01-research-protocols.md         # Clinical trials research and FDA guidance\n├── 02-protocol-foundation.md        # Protocol sections 1-6 (foundation, design, population)\n├── 03-protocol-intervention.md      # Protocol sections 7-8 (intervention details)\n├── 04-protocol-operations.md        # Protocol sections 9-12 (assessments, statistics, operations)\n└── 05-generate-document.md          # NIH Protocol generation\n```\n\n### Utility Scripts\n\n```\nscripts\u002F\n├── sample_size_calculator.py        # Statistical power analysis\n└── test_sample_size_calculator.py   # Unit tests for the calculator\n```\n\n## Prerequisites\n\n### 1. clinical trials MCP Server (Required)\n\n**Installation:**\n- Install via drag-and-drop `.mcpb` file into Claude Desktop\n- Or configure manually in Claude Desktop settings\n\n**Available Tools:**\n`search_clinical_trials` - Search by:\n\ncondition - Disease or condition (e.g., \"pancreatic cancer\")\nintervention - Drug, device, or treatment (e.g., \"pembrolizumab\", \"CAR-T\")\nsponsor - Sponsor or collaborator name (e.g., \"Pfizer\", \"NIH\")\nlocation - City, state, or country (e.g., \"California\", \"Boston\")\nstatus - \"recruiting\" (default), \"active\", \"completed\", \"all\"\nphase - Trial phase: \"1\", \"2\", \"3\", \"4\", \"early_phase1\"\nmax_results - Default 25, max 100\n\n\n`get_trial_details` - Get comprehensive details for a specific trial using its nct_id (e.g., \"NCT04267848\"). Returns eligibility criteria, outcomes, study design, and contact information.\n\n**Verification:** Step 1 will automatically test MCP connectivity at startup.\n\n### 2. FDA Database Access (Built-in)\n\n**Purpose:** FDA regulatory pathway research via explicit database URLs\n\n**Sources:**\n- Step 1: FDA device\u002Fdrug databases (510(k), PMA, De Novo, Drugs@FDA, Orange Book, Purple Book)\n- All sources use direct FDA database URLs - no generic web searches\n\n### 3. Clinical Protocol Template\n\n**Template Files:** Any `.md` files in the `assets\u002F` directory\n\n**Purpose:** Reference template for protocol structure and content guidance. The system automatically detects available templates and uses them dynamically.\n\n### 4. Python Dependencies (Required for Step 2)\n\n**Installation:**\n```bash\npip install -r requirements.txt\n```\n\n**Dependencies:**\n- scipy >= 1.11.0 (statistical calculations)\n- numpy >= 1.24.0 (numerical operations)\n\n**Purpose:** Statistical sample size calculations for clinical protocols\n\n## How to Use\n\nSimply invoke the skill and select your desired mode:\n\n**🔬 Research Only Mode:**\n1. Select \"Research Only\" from the main menu\n2. Provide intervention information\n3. Receive comprehensive research summary as formatted .md artifact\n4. Option to continue with full protocol generation or exit\n\n**📄 Full Protocol Mode:**\n1. Select \"Full Protocol\" from the main menu\n2. Guide you through all steps sequentially (Steps 0-5)\n3. Pause after Step 4 to review the draft protocol\n4. Generate the final protocol document when ready\n\n**Resume Capability:** If interrupted, simply restart the skill and it will automatically resume from your last completed step.\n\n## Execution Flow\n\n### Startup: Welcome and Mode Selection\n\nWhen skill is invoked, display the following message:\n\n```\n🧬 CLINICAL TRIAL PROTOCOL\n\nWelcome! This skill generates clinical trial protocols for medical devices or drugs.\n\n[If waypoints\u002Fintervention_metadata.json exists:]\n✓ Found existing protocol in progress: [Intervention Name]\n  Type: [Device\u002FDrug]\n  Completed: [List of completed steps]\n  Next: [Next step to execute]\n\n📋 SELECT MODE:\n\n1. 🔬 Research Only - Run clinical research analysis (Steps 0-1)\n   • Collect intervention information\n   • Research similar clinical trials\n   • Find FDA guidance and regulatory pathways\n   • Generate comprehensive research summary as .md artifact\n\n2. 📄 Full Protocol - Generate complete clinical trial protocol (Steps 0-5)\n   • Everything in Research Only, plus:\n   • Generate all protocol sections\n   • Create professional protocol document\n\n3. ❌ Exit\n\nPlease select an option (1, 2, or 3):\n```\n\n**🛑 STOP and WAIT for user selection (1, 2, or 3)**\n\n- If **1 (Research Only)**: Set `execution_mode = \"research_only\"` and proceed to Research Only Workflow Logic\n- If **2 (Full Protocol)**: Set `execution_mode = \"full_protocol\"` and proceed to Full Workflow Logic\n- If **3 (Exit)**: Exit gracefully with \"No problem! Restart the skill anytime to continue.\"\n\n---\n\n### Research Only Workflow Logic\n\n**This workflow executes only Steps 0 and 1, then generates a formatted research summary artifact.**\n\n**Step 1: Check for Existing Waypoints**\n- If `waypoints\u002Fintervention_metadata.json` exists: Load metadata, check if steps 0 and 1 are already complete\n- If no metadata exists: Start from Step 0\n\n**Step 2: Execute Research Steps (0 and 1)**\n\nFor each step (0, 1):\n\n1. **Check completion status:** If step already completed in metadata, skip with \"✓ Step [X] already complete\"\n\n2. **Execute step:**\n   - Display \"▶ Executing Step [X]...\"\n   - Read and follow the corresponding subskill file instructions\n   - Wait for completion\n   - Display \"✓ Step [X] complete\"\n   - **Step execution method (ON-DEMAND LOADING):** When a step is ready to execute (NOT before), read the subskill markdown file and execute ALL instructions within it\n   - **Step-to-file mapping:**\n     - Step 0: `references\u002F00-initialize-intervention.md` (collect intervention info)\n     - Step 1: `references\u002F01-research-protocols.md` (clinical research and FDA guidance)\n\n3. **Handle errors:** If step fails, ask user to retry or exit. Save current state for resume capability.\n\n**Step 3: Generate Research Summary Artifact**\n\nAfter Step 1 completes successfully:\n\n1. **Read waypoint files:**\n   - `waypoints\u002Fintervention_metadata.json` (intervention details)\n   - `waypoints\u002F01_clinical_research_summary.json` (research findings)\n\n2. **Create formatted markdown summary:** Generate a comprehensive, well-formatted research summary as a markdown artifact with the following structure:\n\n```markdown\n# Clinical Research Summary: [Intervention Name]\n\n## Intervention Overview\n- **Type:** [Device\u002FDrug]\n- **Indication:** [Target condition\u002Fdisease]\n- **Description:** [Brief intervention description]\n- **Mechanism of Action:** [How it works]\n\n## Similar Clinical Trials\n[List top 5-10 similar trials with NCT ID, title, phase, status, key findings]\n\n## FDA Regulatory Pathway\n- **Recommended Pathway:** [510(k), PMA, De Novo, IND, NDA, BLA, etc.]\n- **Regulatory Basis:** [Rationale for pathway selection]\n- **Key Requirements:** [Major regulatory considerations]\n\n## FDA Guidance Documents\n[List relevant FDA guidance documents with links and key excerpts]\n\n## Study Design Recommendations\n- **Suggested Study Type:** [RCT, single-arm, etc.]\n- **Phase Recommendation:** [Phase 1, 2, 3, etc.]\n- **Primary Endpoint Suggestions:** [Based on similar trials]\n- **Sample Size Considerations:** [Preliminary thoughts]\n\n## Key Insights and Recommendations\n[Synthesized recommendations for protocol development]\n\n## Next Steps\n[If user wants to proceed with full protocol development]\n\n---\n*Generated by Clinical Trial Protocol Skill*\n*Date: [Current date]*\n```\n\n3. **Save artifact:** Write the formatted summary to `waypoints\u002Fresearch_summary.md`\n\n4. **Display completion message:**\n\n```\n✅ RESEARCH COMPLETE\n\nResearch Summary Generated: waypoints\u002Fresearch_summary.md\n\n📊 Key Findings:\n  • Similar Trials Found: [X trials]\n  • Recommended Pathway: [Pathway name]\n  • FDA Guidance Documents: [X documents identified]\n  • Study Design: [Recommended design]\n\n📄 The research summary has been saved as a formatted markdown artifact.\n\nWould you like to:\n1. Continue with full protocol generation (steps 2-5)\n2. Exit and review research summary\n\n```\n\n**Option 1 Logic (Continue to Full Protocol):**\n- Set `execution_mode = \"full_protocol\"`\n- Continue to Full Workflow Logic starting from Step 2 (since 0 and 1 are complete)\n\n**Option 2 Logic (Exit):**\n- Display: \"✓ Research summary saved. Restart the skill anytime to continue with protocol generation.\"\n- Exit orchestrator gracefully\n\n---\n\n### Full Workflow Logic\n\n**Step 1: Check for Existing Waypoints**\n- If `waypoints\u002Fintervention_metadata.json` exists: Load metadata, check `completed_steps` array, resume from next incomplete step\n- If no metadata exists: Start from Step 0\n\n**Step 2: Execute Steps in Order**\n\nFor each step (0, 1, 2, 3, 4, 5):\n\n1. **Check completion status:** If step already completed in metadata, skip with \"✓ Step [X] already complete\"\n\n2. **Execute step:** Display \"▶ Executing Step [X]...\", read and follow the corresponding subskill file instructions, wait for completion, display \"✓ Step [X] complete\"\n   - **Step execution method (ON-DEMAND LOADING):** When a step is ready to execute (NOT before), read the subskill markdown file and execute ALL instructions within it\n   - **IMPORTANT:** Do NOT read subskill files in advance. Only read them at the moment of execution.\n   - **Step-to-file mapping:**\n     - Step 0: `references\u002F00-initialize-intervention.md` (read when Step 0 executes)\n     - Step 1: `references\u002F01-research-protocols.md` (read when Step 1 executes)\n     - Step 2: `references\u002F02-protocol-foundation.md` (read when Step 2 executes - sections 1-6)\n     - Step 3: `references\u002F03-protocol-intervention.md` (read when Step 3 executes - sections 7-8)\n     - Step 4: `references\u002F04-protocol-operations.md` (read when Step 4 executes - sections 9-12)\n     - Step 5: `references\u002F05-concatenate-protocol.md` (read when Step 5 executes - final concatenation)\n\n3. **Handle errors:** If step fails, ask user to retry or exit. Save current state for resume capability.\n\n4. **Display progress:** \"Progress: [X\u002F6] steps complete\", show estimated remaining time\n\n5. **Step 4 Completion Pause:** After Step 4 completes, pause and display the Protocol Completion Menu (see below). Wait for user selection before proceeding.\n\n**Step 2.5: Protocol Completion Menu**\n\nAfter Step 4 completes successfully, display the EXACT menu below (do not improvise or create alternative options):\n\n```\n✅ PROTOCOL COMPLETE: Protocol Draft Generated\n\nProtocol Details:\n  • Study Design: [Design from metadata]\n  • Sample Size: [N subjects from metadata]\n  • Primary Endpoint: [Endpoint from metadata]\n  • Study Duration: [Duration from metadata]\n\nProtocol file: waypoints\u002F02_protocol_draft.md\nFile size: [Size in KB]\n\n📋 WHAT WOULD YOU LIKE TO DO NEXT?\n\n1. 📄 Review Protocol in Artifact - click on the .md file above\n\n2. 📄 Concatenate Final Protocol (Step 5)\n\n3. ⏸️  Exit and Review Later\n\n```\n\n**Option 1 Logic (Review in Artifact):**\nPause, let user open the section files, wait for further instruction\n\n**Option 2 Logic (Concatenate Protocol):**\n1. Execute Step 5 by reading and following `references\u002F05-concatenate-protocol.md`\n2. Step 5 will concatenate all section files into final protocol document\n3. Continue to Step 3 (Final Summary) after Step 5 completes\n\n**Option 3 Logic (Exit):**\n1. Display: \"✓ Protocol sections saved. You can resume with Step 5 anytime to concatenate.\"\n2. Exit orchestrator gracefully\n\n**Step 3: Final Summary**\n\nDisplay completion message with:\n- Intervention name, type (device\u002Fdrug), indication\n- Protocol details (design, sample size, endpoints, duration)\n- All completed steps list\n- Final deliverable: Complete protocol markdown file location (waypoints\u002Fprotocol_complete.md)\n- Waypoint files list for reference\n- Important disclaimers (FDA Pre-Sub, biostatistician review, IRB approval required)\n- Thank you message\n\n## Technical Details\n\n### Waypoint File Formats\n\n**JSON Waypoints** (Steps 0, 1):\n- Structured data for programmatic access\n- Small file sizes (1-15KB)\n- Easy to parse and reference\n\n**Markdown Waypoints** (Steps 2, 3, 4):\n- Step 2: `02_protocol_foundation.md` (Sections 1-6)\n- Step 3: `03_protocol_intervention.md` (Sections 7-8)\n- Step 4: `04_protocol_operations.md` (Sections 9-12)\n- Step 4: `02_protocol_draft.md` (concatenated complete protocol)\n- Human-readable protocol documents\n- Can be directly edited by users\n- Individual section files preserved for easier regeneration\n\n### Data Minimization Strategy\n\nEach step implements aggressive summarization:\n- **Keep:** Top-N results (5-10 max)\n- **Keep:** Key facts and IDs (NCT numbers, endpoint types)\n- **Keep:** Concise rationale (2-3 sentences)\n- **Discard:** Raw MCP query results (not needed after analysis)\n- **Discard:** Full FDA guidance text (only excerpts\u002Fcitations kept)\n- **Discard:** Lower-ranked search results\n\n### Step Independence\n\nEach subskill is designed to:\n- Read only from waypoint files (not conversation history)\n- Produce complete output in single execution\n- Not depend on conversation context from previous steps\n- Be runnable standalone\n\n## Error Handling\n\n### MCP Server Unavailable\n- Detected in: Step 1\n- Action: Display error with installation instructions\n- Allow user to retry after installing MCP server\n- No fallback available - MCP server is required for protocol research\n\n### Step Fails or Returns Error\n- Action: Display error message from subskill\n- Ask user: \"Retry step? (Yes\u002FNo)\"\n  - Yes: Re-run step\n  - No: Save current state, exit orchestrator\n\n### User Interruption\n- All progress saved in waypoint files\n- User can resume anytime by restarting the skill\n- Workflow automatically detects completed steps and resumes from next step\n- No data loss\n\n## Disclaimers\n\n⚠️ **IMPORTANT:** This protocol generation tool provides preliminary clinical study protocol based on NIH\u002FFDA guidelines and similar trials. It does NOT constitute:\n- Official FDA or IRB determination or approval\n- Medical, legal, or regulatory advice\n- Substitute for professional biostatistician review\n- Substitute for FDA Pre-Submission meeting\n- Guarantee of regulatory or clinical success\n\n**REQUIRED before proceeding with clinical study:**\n- Biostatistician review and sample size validation\n- FDA Pre-Submission meeting (Q-Submission for devices, Pre-IND for drugs)\n- IRB review and approval\n- Clinical expert and regulatory consultant engagement\n- Legal review of protocol and informed consent\n- Site investigator review and input\n- Sponsor completion of all [TBD] items in protocol\n\n**PROFESSIONAL CONSULTATION STRONGLY RECOMMENDED**\n\nClinical trial protocols are complex, high-stakes documents requiring expertise across multiple disciplines. Professional consultation with clinical trial experts, biostatisticians, and regulatory affairs specialists is essential before proceeding with clinical study planning.\n\n\n## Implementation Requirements\n\nWhen this skill is invoked:\n\n1. **Display the welcome message with mode selection** (shown in \"Startup: Welcome and Mode Selection\" section)\n\n2. **Wait for user mode selection** (1: Research Only, 2: Full Protocol, 3: Exit)\n\n3. **Execute based on selected mode:**\n   - **Research Only Mode (Option 1):**\n     - Execute Research Only Workflow Logic (Steps 0-1 only)\n     - Generate formatted research summary as .md artifact\n     - Offer option to continue with full protocol or exit\n   - **Full Protocol Mode (Option 2):**\n     - Execute Full Workflow Logic (Steps 0-5)\n     - Check for existing waypoints and resume from last completed step\n     - OR start from Step 0 if no waypoints exist\n     - Execute all steps sequentially until complete\n\n4. **For each step execution (LAZY LOADING - On-Demand Only):**\n   - **ONLY when a step is ready to execute**, read the corresponding subskill file\n   - Do NOT read subskill files in advance or \"to prepare\"\n   - Example: When Step 1 needs to run, THEN read `references\u002F01-research-protocols.md` and follow its instructions\n   - **For protocol development:** Execute Steps 2, 3, 4 sequentially in order\n   - Do NOT try to execute multiple steps in parallel - run sequentially\n   - Read each step's subskill file only when that specific step is about to execute\n\n5. **Research summary artifact generation (Research Only Mode):**\n   - After Step 1 completes, read waypoint files\n   - Generate comprehensive, well-formatted markdown summary\n   - Save to `waypoints\u002Fresearch_summary.md`\n   - Display completion message with key findings\n\n6. **Handle errors gracefully:**\n   - If a step fails, give user option to retry or exit\n   - If MCP server unavailable, explain how to install\n   - All progress is saved automatically in waypoints\n\n7. **Track progress:**\n   - Update `waypoints\u002Fintervention_metadata.json` after each step\n   - Show progress indicators to user (e.g., \"Progress: 3\u002F6 steps complete\" or \"Progress: 2\u002F2 research steps complete\")\n   - Provide clear feedback on what's happening\n\n8. **Final output:**\n   - **Research Only:** Display research summary location and offer to continue with full protocol\n   - **Full Protocol:** Congratulate user, display protocol location and next steps\n   - Remind user of disclaimers\n\n\n",{"name":5,"description":7},"main","0d38fe5c0a50d711ebe6ecd85dc84b98660579d0",115,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences","https:\u002F\u002Fgithub.com\u002Fanthropics","https:\u002F\u002Fanthropic.com","2026-11-09 20:22:16",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add anthropics\u002Flife-sciences --skill clinical-trial-protocol-skill",{"tool":38,"command":39},"bunx","bunx skills add anthropics\u002Flife-sciences --skill clinical-trial-protocol-skill",{"tool":41,"command":42},"pnpm","pnpm dlx skills add anthropics\u002Flife-sciences --skill clinical-trial-protocol-skill","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002FSKILL.md",true,[46,52,57,63,67,71,75,79,83,87,93],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"assets\u002FFDA-Clinical-Protocol-Template.md","ASSET","资产文件",176697,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002Fassets\u002FFDA-Clinical-Protocol-Template.md",{"path":53,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":56,"sourceHtmlUrl":16},"LICENSE.txt","LICENSE","许可证",10254,{"path":58,"resourceType":59,"resourceTypeDesc":60,"sizeBytes":61,"sourceHtmlUrl":62},"README.md","REFERENCE","参考文档",2967,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002FREADME.md",{"path":64,"resourceType":59,"resourceTypeDesc":60,"sizeBytes":65,"sourceHtmlUrl":66},"references\u002F00-initialize-intervention.md",8823,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002Freferences\u002F00-initialize-intervention.md",{"path":68,"resourceType":59,"resourceTypeDesc":60,"sizeBytes":69,"sourceHtmlUrl":70},"references\u002F01-research-protocols.md",19123,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002Freferences\u002F01-research-protocols.md",{"path":72,"resourceType":59,"resourceTypeDesc":60,"sizeBytes":73,"sourceHtmlUrl":74},"references\u002F02-protocol-foundation.md",40691,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002Freferences\u002F02-protocol-foundation.md",{"path":76,"resourceType":59,"resourceTypeDesc":60,"sizeBytes":77,"sourceHtmlUrl":78},"references\u002F03-protocol-intervention.md",33535,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002Freferences\u002F03-protocol-intervention.md",{"path":80,"resourceType":59,"resourceTypeDesc":60,"sizeBytes":81,"sourceHtmlUrl":82},"references\u002F04-protocol-operations.md",84394,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002Freferences\u002F04-protocol-operations.md",{"path":84,"resourceType":59,"resourceTypeDesc":60,"sizeBytes":85,"sourceHtmlUrl":86},"references\u002F05-concatenate-protocol.md",8998,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002Freferences\u002F05-concatenate-protocol.md",{"path":88,"resourceType":89,"resourceTypeDesc":90,"sizeBytes":91,"sourceHtmlUrl":92},"scripts\u002Fsample_size_calculator.py","SCRIPT","脚本",16265,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002Fscripts\u002Fsample_size_calculator.py",{"path":94,"resourceType":89,"resourceTypeDesc":90,"sizeBytes":95,"sourceHtmlUrl":96},"scripts\u002Ftest_sample_size_calculator.py",5541,"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-protocol-skill\u002Fscripts\u002Ftest_sample_size_calculator.py",[98,106,114],{"skillNo":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":102,"sourceHtmlUrl":103,"categories":104,"keywords":105,"publishedAt":20},"01M73FDQSZBB1YJ13MGR2E0CTN","clinical-trial-forecast","Forecasts one registered clinical trial's probability of success from registry records and documents the user supplies: checked research, a critiqued Monte Carlo model, a cited report and a claim ledger. Runs in Claude Science with Delegation on. Use when the user asks for a trial's probability of success, asks whether a named trial (an NCT id) will meet its primary endpoint, or asks to run the clinical-trial-forecast skill.","https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-forecast\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fclinical-trial-forecast\u002FSKILL.md",[],[],{"skillNo":107,"name":108,"nameZh":6,"description":109,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":110,"sourceHtmlUrl":111,"categories":112,"keywords":113,"publishedAt":20},"01M73FDGTXZKH9KWAM8XQCC66B","binder-gauntlet","Validation and developability gauntlet for de novo designed protein binders and miniproteins against any protein target. Use this skill when designed binders are scored only by structure-predictor confidence (ipSAE, ipTM, pAE) and you must validate binding with independent re-prediction, pose agreement (DockQ) and controls, profile developability and sequence liabilities, check specificity and cross-reactivity, and write a validation dossier before you rank, filter or finalise binder designs, and at every selection round of a design campaign to steer it. Use when a user says \"validate these binder designs\", \"run the gauntlet on these\", \"check developability\" or \"sequence liabilities\", asks which designs should go to the lab, or asks for controls, a specificity panel or a validation dossier for designed binders.","https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fbinder-gauntlet\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fbinder-gauntlet\u002FSKILL.md",[],[],{"skillNo":115,"name":116,"nameZh":6,"description":117,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":118,"sourceHtmlUrl":119,"categories":120,"keywords":126,"publishedAt":147},"01M718EB79N3YZ3R11X0CVBCG6","scientific-problem-selection","This skill should be used when scientists need help with research problem selection, project ideation, troubleshooting stuck projects, or strategic scientific decisions. Use this skill when users ask to pitch a new research idea, work through a project problem, evaluate project risks, plan research strategy, navigate decision trees, or get help choosing what scientific problem to work on. Typical requests include \"I have an idea for a project\", \"I'm stuck on my research\", \"help me evaluate this project\", \"what should I work on\", or \"I need strategic advice about my research\".","https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fscientific-problem-selection\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fanthropics\u002Flife-sciences\u002Fblob\u002Fmain\u002Fscientific-problem-selection\u002FSKILL.md",[121],{"categoryNo":122,"categoryName":123,"taxonomyType":124,"taxonomyTypeDesc":125},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146],"selection","scientific","scientific problem selection","problem","idea","my","me","plan","work","navigate","ask","when","pitch","new","Research Strategy","decision","work on","project","get","projects","2026-11-10 10:12:30"]