[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M73G671KC62KXBDX96TM9753":3,"skill-same:01M73G671KC62KXBDX96TM9753":49},{"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":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":25,"commitSha":26,"skillDirectory":27,"forksCount":28,"repositoryHtmlUrl":29,"repositoryHomepageUrl":30,"ownerHtmlUrl":31,"ownerBlogUrl":32,"ownerDescription":33,"updatedAt":34,"install":35,"resources":48},"01M73G671KC62KXBDX96TM9753","create-meta-prompts",null,"Create optimized prompts for Claude-to-Claude pipelines with research, planning, and execution stages. Use when building prompts that produce outputs for other prompts to consume, or when running multi-stage workflows (research -> plan -> implement).","Horonomos-laps\u002FYana-AI","Horonomos-laps","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F355299122?v=4","01M76B3QPXE88ET2SQPDPVATZT"," Horonomos-laps","01M72ZQRJPCYPSSWT4D00VJR2A",13,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002Fcore\u002Fskills\u002Ftaches--create-meta-prompts\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: create-meta-prompts\ndescription: Create optimized prompts for Claude-to-Claude pipelines with research, planning, and execution stages. Use when building prompts that produce outputs for other prompts to consume, or when running multi-stage workflows (research -> plan -> implement).\norigin: \"https:\u002F\u002Fraw.githubusercontent.com\u002Fglittercowboy\u002Ftaches-cc-resources\u002Fmain\u002Fskills\u002Fcreate-meta-prompts\u002FSKILL.md\"\n---\n\n\u003Cobjective>\nCreate prompts optimized for Claude-to-Claude communication in multi-stage workflows. Outputs are structured with XML and metadata for efficient parsing by subsequent prompts.\n\nEvery execution produces a `SUMMARY.md` for quick human scanning without reading full outputs.\n\nEach prompt gets its own folder in `.prompts\u002F` with its output artifacts, enabling clear provenance and chain detection.\n\u003C\u002Fobjective>\n\n\u003Cquick_start>\n\u003Cworkflow>\n1. **Intake**: Determine purpose (Do\u002FPlan\u002FResearch\u002FRefine), gather requirements\n2. **Chain detection**: Check for existing research\u002Fplan files to reference\n3. **Generate**: Create prompt using purpose-specific patterns\n4. **Save**: Create folder in `.prompts\u002F{number}-{topic}-{purpose}\u002F`\n5. **Present**: Show decision tree for running\n6. **Execute**: Run prompt(s) with dependency-aware execution engine\n7. **Summarize**: Create SUMMARY.md for human scanning\n\u003C\u002Fworkflow>\n\n\u003Cfolder_structure>\n```\n.prompts\u002F\n├── 001-auth-research\u002F\n│   ├── completed\u002F\n│   │   └── 001-auth-research.md    # Prompt (archived after run)\n│   ├── auth-research.md            # Full output (XML for Claude)\n│   └── SUMMARY.md                  # Executive summary (markdown for human)\n├── 002-auth-plan\u002F\n│   ├── completed\u002F\n│   │   └── 002-auth-plan.md\n│   ├── auth-plan.md\n│   └── SUMMARY.md\n├── 003-auth-implement\u002F\n│   ├── completed\u002F\n│   │   └── 003-auth-implement.md\n│   └── SUMMARY.md                  # Do prompts create code elsewhere\n├── 004-auth-research-refine\u002F\n│   ├── completed\u002F\n│   │   └── 004-auth-research-refine.md\n│   ├── archive\u002F\n│   │   └── auth-research-v1.md     # Previous version\n│   └── SUMMARY.md\n```\n\u003C\u002Ffolder_structure>\n\u003C\u002Fquick_start>\n\n\u003Ccontext>\nPrompts directory: !`[ -d .\u002F.prompts ] && echo \"exists\" || echo \"missing\"`\nExisting research\u002Fplans: !`find .\u002F.prompts -name \"*-research.md\" -o -name \"*-plan.md\" 2>\u002Fdev\u002Fnull | head -10`\nNext prompt number: !`ls -d .\u002F.prompts\u002F*\u002F 2>\u002Fdev\u002Fnull | wc -l | xargs -I {} expr {} + 1`\n\u003C\u002Fcontext>\n\n\u003Cautomated_workflow>\n\n\u003Cstep_0_intake_gate>\n\u003Ctitle>Adaptive Requirements Gathering\u003C\u002Ftitle>\n\n\u003Ccritical_first_action>\n**BEFORE analyzing anything**, check if context was provided.\n\nIF no context provided (skill invoked without description):\n→ **IMMEDIATELY use AskUserQuestion** with:\n\n- header: \"Purpose\"\n- question: \"What is the purpose of this prompt?\"\n- options:\n  - \"Do\" - Execute a task, produce an artifact\n  - \"Plan\" - Create an approach, roadmap, or strategy\n  - \"Research\" - Gather information or understand something\n  - \"Refine\" - Improve an existing research or plan output\n\nAfter selection, ask: \"Describe what you want to accomplish\" (they select \"Other\" to provide free text).\n\nIF context was provided:\n→ Check if purpose is inferable from keywords:\n  - `implement`, `build`, `create`, `fix`, `add`, `refactor` → Do\n  - `plan`, `roadmap`, `approach`, `strategy`, `decide`, `phases` → Plan\n  - `research`, `understand`, `learn`, `gather`, `analyze`, `explore` → Research\n  - `refine`, `improve`, `deepen`, `expand`, `iterate`, `update` → Refine\n\n→ If unclear, ask the Purpose question above as first contextual question\n→ If clear, proceed to adaptive_analysis with inferred purpose\n\u003C\u002Fcritical_first_action>\n\n\u003Cadaptive_analysis>\nExtract and infer:\n\n- **Purpose**: Do, Plan, Research, or Refine\n- **Topic identifier**: Kebab-case identifier for file naming (e.g., `auth`, `stripe-payments`)\n- **Complexity**: Simple vs complex (affects prompt depth)\n- **Prompt structure**: Single vs multiple prompts\n- **Target** (Refine only): Which existing output to improve\n\nIf topic identifier not obvious, ask:\n- header: \"Topic\"\n- question: \"What topic\u002Ffeature is this for? (used for file naming)\"\n- Let user provide via \"Other\" option\n- Enforce kebab-case (convert spaces\u002Funderscores to hyphens)\n\nFor Refine purpose, also identify target output from `.prompts\u002F*\u002F` to improve.\n\u003C\u002Fadaptive_analysis>\n\n\u003Cchain_detection>\nScan `.prompts\u002F*\u002F` for existing `*-research.md` and `*-plan.md` files.\n\nIf found:\n1. List them: \"Found existing files: auth-research.md (in 001-auth-research\u002F), stripe-plan.md (in 005-stripe-plan\u002F)\"\n2. Use AskUserQuestion:\n   - header: \"Reference\"\n   - question: \"Should this prompt reference any existing research or plans?\"\n   - options: List found files + \"None\"\n   - multiSelect: true\n\nMatch by topic keyword when possible (e.g., \"auth plan\" → suggest auth-research.md).\n\u003C\u002Fchain_detection>\n\n\u003Ccontextual_questioning>\nGenerate 2-4 questions using AskUserQuestion based on purpose and gaps.\n\nLoad questions from: [references\u002Fquestion-bank.md](references\u002Fquestion-bank.md)\n\nRoute by purpose:\n- Do → artifact type, scope, approach\n- Plan → plan purpose, format, constraints\n- Research → depth, sources, output format\n- Refine → target selection, feedback, preservation\n\u003C\u002Fcontextual_questioning>\n\n\u003Cdecision_gate>\nAfter receiving answers, present decision gate using AskUserQuestion:\n\n- header: \"Ready\"\n- question: \"Ready to create the prompt?\"\n- options:\n  - \"Proceed\" - Create the prompt with current context\n  - \"Ask more questions\" - I have more details to clarify\n  - \"Let me add context\" - I want to provide additional information\n\nLoop until \"Proceed\" selected.\n\u003C\u002Fdecision_gate>\n\n\u003Cfinalization>\nAfter \"Proceed\" selected, state confirmation:\n\n\"Creating a {purpose} prompt for: {topic}\nFolder: .prompts\u002F{number}-{topic}-{purpose}\u002F\nReferences: {list any chained files}\"\n\nThen proceed to generation.\n\u003C\u002Ffinalization>\n\u003C\u002Fstep_0_intake_gate>\n\n\u003Cstep_1_generate>\n\u003Ctitle>Generate Prompt\u003C\u002Ftitle>\n\nLoad purpose-specific patterns:\n- Do: [references\u002Fdo-patterns.md](references\u002Fdo-patterns.md)\n- Plan: [references\u002Fplan-patterns.md](references\u002Fplan-patterns.md)\n- Research: [references\u002Fresearch-patterns.md](references\u002Fresearch-patterns.md)\n- Refine: [references\u002Frefine-patterns.md](references\u002Frefine-patterns.md)\n\nLoad intelligence rules: [references\u002Fintelligence-rules.md](references\u002Fintelligence-rules.md)\n\n\u003Cprompt_structure>\nAll generated prompts include:\n\n1. **Objective**: What to accomplish, why it matters\n2. **Context**: Referenced files (@), dynamic context (!)\n3. **Requirements**: Specific instructions for the task\n4. **Output specification**: Where to save, what structure\n5. **Metadata requirements**: For research\u002Fplan outputs, specify XML metadata structure\n6. **SUMMARY.md requirement**: All prompts must create a SUMMARY.md file\n7. **Success criteria**: How to know it worked\n\nFor Research and Plan prompts, output must include:\n- `\u003Cconfidence>` - How confident in findings\n- `\u003Cdependencies>` - What's needed to proceed\n- `\u003Copen_questions>` - What remains uncertain\n- `\u003Cassumptions>` - What was assumed\n\nAll prompts must create `SUMMARY.md` with:\n- **One-liner** - Substantive description of outcome\n- **Version** - v1 or iteration info\n- **Key Findings** - Actionable takeaways\n- **Files Created** - (Do prompts only)\n- **Decisions Needed** - What requires user input\n- **Blockers** - External impediments\n- **Next Step** - Concrete forward action\n\u003C\u002Fprompt_structure>\n\n\u003Cfile_creation>\n1. Create folder: `.prompts\u002F{number}-{topic}-{purpose}\u002F`\n2. Create `completed\u002F` subfolder\n3. Write prompt to: `.prompts\u002F{number}-{topic}-{purpose}\u002F{number}-{topic}-{purpose}.md`\n4. Prompt instructs output to: `.prompts\u002F{number}-{topic}-{purpose}\u002F{topic}-{purpose}.md`\n\u003C\u002Ffile_creation>\n\u003C\u002Fstep_1_generate>\n\n\u003Cstep_2_present>\n\u003Ctitle>Present Decision Tree\u003C\u002Ftitle>\n\nAfter saving prompt(s), present inline (not AskUserQuestion):\n\n\u003Csingle_prompt_presentation>\n```\nPrompt created: .prompts\u002F{number}-{topic}-{purpose}\u002F{number}-{topic}-{purpose}.md\n\nWhat's next?\n\n1. Run prompt now\n2. Review\u002Fedit prompt first\n3. Save for later\n4. Other\n\nChoose (1-4): _\n```\n\u003C\u002Fsingle_prompt_presentation>\n\n\u003Cmulti_prompt_presentation>\n```\nPrompts created:\n- .prompts\u002F001-auth-research\u002F001-auth-research.md\n- .prompts\u002F002-auth-plan\u002F002-auth-plan.md\n- .prompts\u002F003-auth-implement\u002F003-auth-implement.md\n\nDetected execution order: Sequential (002 references 001 output, 003 references 002 output)\n\nWhat's next?\n\n1. Run all prompts (sequential)\n2. Review\u002Fedit prompts first\n3. Save for later\n4. Other\n\nChoose (1-4): _\n```\n\u003C\u002Fmulti_prompt_presentation>\n\u003C\u002Fstep_2_present>\n\n\u003Cstep_3_execute>\n\u003Ctitle>Execution Engine\u003C\u002Ftitle>\n\n\u003Cexecution_modes>\n\u003Csingle_prompt>\nStraightforward execution of one prompt.\n\n1. Read prompt file contents\n2. Spawn Task agent with subagent_type=\"general-purpose\"\n3. Include in task prompt:\n   - The complete prompt contents\n   - Output location: `.prompts\u002F{number}-{topic}-{purpose}\u002F{topic}-{purpose}.md`\n4. Wait for completion\n5. Validate output (see validation section)\n6. Archive prompt to `completed\u002F` subfolder\n7. Report results with next-step options\n\u003C\u002Fsingle_prompt>\n\n\u003Csequential_execution>\nFor chained prompts where each depends on previous output.\n\n1. Build execution queue from dependency order\n2. For each prompt in queue:\n   a. Read prompt file\n   b. Spawn Task agent\n   c. Wait for completion\n   d. Validate output\n   e. If validation fails → stop, report failure, offer recovery options\n   f. If success → archive prompt, continue to next\n3. Report consolidated results\n\n\u003Cprogress_reporting>\nShow progress during execution:\n```\nExecuting 1\u002F3: 001-auth-research... ✓\nExecuting 2\u002F3: 002-auth-plan... ✓\nExecuting 3\u002F3: 003-auth-implement... (running)\n```\n\u003C\u002Fprogress_reporting>\n\u003C\u002Fsequential_execution>\n\n\u003Cparallel_execution>\nFor independent prompts with no dependencies.\n\n1. Read all prompt files\n2. **CRITICAL**: Spawn ALL Task agents in a SINGLE message\n   - This is required for true parallel execution\n   - Each task includes its output location\n3. Wait for all to complete\n4. Validate all outputs\n5. Archive all prompts\n6. Report consolidated results (successes and failures)\n\n\u003Cfailure_handling>\nUnlike sequential, parallel continues even if some fail:\n- Collect all results\n- Archive successful prompts\n- Report failures with details\n- Offer to retry failed prompts\n\u003C\u002Ffailure_handling>\n\u003C\u002Fparallel_execution>\n\n\u003Cmixed_dependencies>\nFor complex DAGs (e.g., two parallel research → one plan).\n\n1. Analyze dependency graph from @ references\n2. Group into execution layers:\n   - Layer 1: No dependencies (run parallel)\n   - Layer 2: Depends only on layer 1 (run after layer 1 completes)\n   - Layer 3: Depends on layer 2, etc.\n3. Execute each layer:\n   - Parallel within layer\n   - Sequential between layers\n4. Stop if any dependency fails (downstream prompts can't run)\n\n\u003Cexample>\n```\nLayer 1 (parallel): 001-api-research, 002-db-research\nLayer 2 (after layer 1): 003-architecture-plan\nLayer 3 (after layer 2): 004-implement\n```\n\u003C\u002Fexample>\n\u003C\u002Fmixed_dependencies>\n\u003C\u002Fexecution_modes>\n\n\u003Cdependency_detection>\n\u003Cautomatic_detection>\nScan prompt contents for @ references to determine dependencies:\n\n1. Parse each prompt for `@.prompts\u002F{number}-{topic}\u002F` patterns\n2. Build dependency graph\n3. Detect cycles (error if found)\n4. Determine execution order\n\n\u003Cinference_rules>\nIf no explicit @ references found, infer from purpose:\n- Research prompts: No dependencies (can parallel)\n- Plan prompts: Depend on same-topic research\n- Do prompts: Depend on same-topic plan\n\nOverride with explicit references when present.\n\u003C\u002Finference_rules>\n\u003C\u002Fautomatic_detection>\n\n\u003Cmissing_dependencies>\nIf a prompt references output that doesn't exist:\n\n1. Check if it's another prompt in this session (will be created)\n2. Check if it exists in `.prompts\u002F*\u002F` (already completed)\n3. If truly missing:\n   - Warn user: \"002-auth-plan references auth-research.md which doesn't exist\"\n   - Offer: Create the missing research prompt first? \u002F Continue anyway? \u002F Cancel?\n\u003C\u002Fmissing_dependencies>\n\u003C\u002Fdependency_detection>\n\n\u003Cvalidation>\n\u003Coutput_validation>\nAfter each prompt completes, verify success:\n\n1. **File exists**: Check output file was created\n2. **Not empty**: File has content (> 100 chars)\n3. **Metadata present** (for research\u002Fplan): Check for required XML tags\n   - `\u003Cconfidence>`\n   - `\u003Cdependencies>`\n   - `\u003Copen_questions>`\n   - `\u003Cassumptions>`\n4. **SUMMARY.md exists**: Check SUMMARY.md was created\n5. **SUMMARY.md complete**: Has required sections (Key Findings, Decisions Needed, Blockers, Next Step)\n6. **One-liner is substantive**: Not generic like \"Research completed\"\n\n\u003Cvalidation_failure>\nIf validation fails:\n- Report what's missing\n- Offer options:\n  - Retry the prompt\n  - Continue anyway (for non-critical issues)\n  - Stop and investigate\n\u003C\u002Fvalidation_failure>\n\u003C\u002Foutput_validation>\n\u003C\u002Fvalidation>\n\n\u003Cfailure_handling>\n\u003Csequential_failure>\nStop the chain immediately:\n```\n✗ Failed at 2\u002F3: 002-auth-plan\n\nCompleted:\n- 001-auth-research ✓ (archived)\n\nFailed:\n- 002-auth-plan: Output file not created\n\nNot started:\n- 003-auth-implement\n\nWhat's next?\n1. Retry 002-auth-plan\n2. View error details\n3. Stop here (keep completed work)\n4. Other\n```\n\u003C\u002Fsequential_failure>\n\n\u003Cparallel_failure>\nContinue others, report all results:\n```\nParallel execution completed with errors:\n\n✓ 001-api-research (archived)\n✗ 002-db-research: Validation failed - missing \u003Cconfidence> tag\n✓ 003-ui-research (archived)\n\nWhat's next?\n1. Retry failed prompt (002)\n2. View error details\n3. Continue without 002\n4. Other\n```\n\u003C\u002Fparallel_failure>\n\u003C\u002Ffailure_handling>\n\n\u003Carchiving>\n\u003Carchive_timing>\n- **Sequential**: Archive each prompt immediately after successful completion\n  - Provides clear state if execution stops mid-chain\n- **Parallel**: Archive all at end after collecting results\n  - Keeps prompts available for potential retry\n\n\u003Carchive_operation>\nMove prompt file to completed subfolder:\n```bash\nmv .prompts\u002F{number}-{topic}-{purpose}\u002F{number}-{topic}-{purpose}.md \\\n   .prompts\u002F{number}-{topic}-{purpose}\u002Fcompleted\u002F\n```\n\nOutput file stays in place (not moved).\n\u003C\u002Farchive_operation>\n\u003C\u002Farchiving>\n\n\u003Cresult_presentation>\n\u003Csingle_result>\n```\n✓ Executed: 001-auth-research\n✓ Created: .prompts\u002F001-auth-research\u002FSUMMARY.md\n\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n# Auth Research Summary\n\n**JWT with jose library and httpOnly cookies recommended**\n\n## Key Findings\n• jose outperforms jsonwebtoken with better TypeScript support\n• httpOnly cookies required (localStorage is XSS vulnerable)\n• Refresh rotation is OWASP standard\n\n## Decisions Needed\nNone - ready for planning\n\n## Blockers\nNone\n\n## Next Step\nCreate auth-plan.md\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\nWhat's next?\n1. Create planning prompt (auth-plan)\n2. View full research output\n3. Done\n4. Other\n```\n\nDisplay the actual SUMMARY.md content inline so user sees findings without opening files.\n\u003C\u002Fsingle_result>\n\n\u003Cchain_result>\n```\n✓ Chain completed: auth workflow\n\nResults:\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n001-auth-research\n**JWT with jose library and httpOnly cookies recommended**\nDecisions: None • Blockers: None\n\n002-auth-plan\n**4-phase implementation: types → JWT core → refresh → tests**\nDecisions: Approve 15-min token expiry • Blockers: None\n\n003-auth-implement\n**JWT middleware complete with 6 files created**\nDecisions: Review before Phase 2 • Blockers: None\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\nAll prompts archived. Full summaries in .prompts\u002F*\u002FSUMMARY.md\n\nWhat's next?\n1. Review implementation\n2. Run tests\n3. Create new prompt chain\n4. Other\n```\n\nFor chains, show condensed one-liner from each SUMMARY.md with decisions\u002Fblockers flagged.\n\u003C\u002Fchain_result>\n\u003C\u002Fresult_presentation>\n\n\u003Cspecial_cases>\n\u003Cre_running_completed>\nIf user wants to re-run an already-completed prompt:\n\n1. Check if prompt is in `completed\u002F` subfolder\n2. Move it back to parent folder\n3. Optionally backup existing output: `{output}.bak`\n4. Execute normally\n\u003C\u002Fre_running_completed>\n\n\u003Coutput_conflicts>\nIf output file already exists:\n\n1. For re-runs: Backup existing → `{filename}.bak`\n2. For new runs: Should not happen (unique numbering)\n3. If conflict detected: Ask user - Overwrite? \u002F Rename? \u002F Cancel?\n\u003C\u002Foutput_conflicts>\n\n\u003Ccommit_handling>\nAfter successful execution:\n\n1. Do NOT auto-commit (user controls git workflow)\n2. Mention what files were created\u002Fmodified\n3. User can commit when ready\n\nException: If user explicitly requests commit, stage and commit:\n- Output files created\n- Prompts archived\n- Any implementation changes (for Do prompts)\n\u003C\u002Fcommit_handling>\n\n\u003Crecursive_prompts>\nIf a prompt's output includes instructions to create more prompts:\n\n1. This is advanced usage - don't auto-detect\n2. Present the output to user\n3. User can invoke skill again to create follow-up prompts\n4. Maintains user control over prompt creation\n\u003C\u002Frecursive_prompts>\n\u003C\u002Fspecial_cases>\n\u003C\u002Fstep_3_execute>\n\n\u003C\u002Fautomated_workflow>\n\n\u003Creference_guides>\n**Prompt patterns by purpose:**\n- [references\u002Fdo-patterns.md](references\u002Fdo-patterns.md) - Execution prompts + output structure\n- [references\u002Fplan-patterns.md](references\u002Fplan-patterns.md) - Planning prompts + plan.md structure\n- [references\u002Fresearch-patterns.md](references\u002Fresearch-patterns.md) - Research prompts + research.md structure\n- [references\u002Frefine-patterns.md](references\u002Frefine-patterns.md) - Iteration prompts + versioning\n\n**Shared templates:**\n- [references\u002Fsummary-template.md](references\u002Fsummary-template.md) - SUMMARY.md structure and field requirements\n- [references\u002Fmetadata-guidelines.md](references\u002Fmetadata-guidelines.md) - Confidence, dependencies, open questions, assumptions\n\n**Supporting references:**\n- [references\u002Fquestion-bank.md](references\u002Fquestion-bank.md) - Intake questions by purpose\n- [references\u002Fintelligence-rules.md](references\u002Fintelligence-rules.md) - Extended thinking, parallel tools, depth decisions\n\u003C\u002Freference_guides>\n\n\u003Csuccess_criteria>\n**Prompt Creation:**\n- Intake gate completed with purpose and topic identified\n- Chain detection performed, relevant files referenced\n- Prompt generated with correct structure for purpose\n- Folder created in `.prompts\u002F` with correct naming\n- Output file location specified in prompt\n- SUMMARY.md requirement included in prompt\n- Metadata requirements included for Research\u002FPlan outputs\n- Quality controls included for Research outputs (verification checklist, QA, pre-submission)\n- Streaming write instructions included for Research outputs\n- Decision tree presented\n\n**Execution (if user chooses to run):**\n- Dependencies correctly detected and ordered\n- Prompts executed in correct order (sequential\u002Fparallel\u002Fmixed)\n- Output validated after each completion\n- SUMMARY.md created with all required sections\n- One-liner is substantive (not generic)\n- Failed prompts handled gracefully with recovery options\n- Successful prompts archived to `completed\u002F` subfolder\n- SUMMARY.md displayed inline in results\n- Results presented with decisions\u002Fblockers flagged\n\n**Research Quality (for Research prompts):**\n- Verification checklist completed\n- Quality report distinguishes verified from assumed claims\n- Sources consulted listed with URLs\n- Confidence levels assigned to findings\n- Critical claims verified with official documentation\n\u003C\u002Fsuccess_criteria>\n","\u003Cobjective>\nCreate prompts optimized for Claude-to-Claude communication in multi-stage workflows. Outputs are structured with XML and metadata for efficient parsing by subsequent prompts.\n\nEvery execution produces a `SUMMARY.md` for quick human scanning without reading full outputs.\n\nEach prompt gets its own folder in `.prompts\u002F` with its output artifacts, enabling clear provenance and chain detection.\n\u003C\u002Fobjective>\n\n\u003Cquick_start>\n\u003Cworkflow>\n1. **Intake**: Determine purpose (Do\u002FPlan\u002FResearch\u002FRefine), gather requirements\n2. **Chain detection**: Check for existing research\u002Fplan files to reference\n3. **Generate**: Create prompt using purpose-specific patterns\n4. **Save**: Create folder in `.prompts\u002F{number}-{topic}-{purpose}\u002F`\n5. **Present**: Show decision tree for running\n6. **Execute**: Run prompt(s) with dependency-aware execution engine\n7. **Summarize**: Create SUMMARY.md for human scanning\n\u003C\u002Fworkflow>\n\n\u003Cfolder_structure>\n```\n.prompts\u002F\n├── 001-auth-research\u002F\n│   ├── completed\u002F\n│   │   └── 001-auth-research.md    # Prompt (archived after run)\n│   ├── auth-research.md            # Full output (XML for Claude)\n│   └── SUMMARY.md                  # Executive summary (markdown for human)\n├── 002-auth-plan\u002F\n│   ├── completed\u002F\n│   │   └── 002-auth-plan.md\n│   ├── auth-plan.md\n│   └── SUMMARY.md\n├── 003-auth-implement\u002F\n│   ├── completed\u002F\n│   │   └── 003-auth-implement.md\n│   └── SUMMARY.md                  # Do prompts create code elsewhere\n├── 004-auth-research-refine\u002F\n│   ├── completed\u002F\n│   │   └── 004-auth-research-refine.md\n│   ├── archive\u002F\n│   │   └── auth-research-v1.md     # Previous version\n│   └── SUMMARY.md\n```\n\u003C\u002Ffolder_structure>\n\u003C\u002Fquick_start>\n\n\u003Ccontext>\nPrompts directory: !`[ -d .\u002F.prompts ] && echo \"exists\" || echo \"missing\"`\nExisting research\u002Fplans: !`find .\u002F.prompts -name \"*-research.md\" -o -name \"*-plan.md\" 2>\u002Fdev\u002Fnull | head -10`\nNext prompt number: !`ls -d .\u002F.prompts\u002F*\u002F 2>\u002Fdev\u002Fnull | wc -l | xargs -I {} expr {} + 1`\n\u003C\u002Fcontext>\n\n\u003Cautomated_workflow>\n\n\u003Cstep_0_intake_gate>\n\u003Ctitle>Adaptive Requirements Gathering\u003C\u002Ftitle>\n\n\u003Ccritical_first_action>\n**BEFORE analyzing anything**, check if context was provided.\n\nIF no context provided (skill invoked without description):\n→ **IMMEDIATELY use AskUserQuestion** with:\n\n- header: \"Purpose\"\n- question: \"What is the purpose of this prompt?\"\n- options:\n  - \"Do\" - Execute a task, produce an artifact\n  - \"Plan\" - Create an approach, roadmap, or strategy\n  - \"Research\" - Gather information or understand something\n  - \"Refine\" - Improve an existing research or plan output\n\nAfter selection, ask: \"Describe what you want to accomplish\" (they select \"Other\" to provide free text).\n\nIF context was provided:\n→ Check if purpose is inferable from keywords:\n  - `implement`, `build`, `create`, `fix`, `add`, `refactor` → Do\n  - `plan`, `roadmap`, `approach`, `strategy`, `decide`, `phases` → Plan\n  - `research`, `understand`, `learn`, `gather`, `analyze`, `explore` → Research\n  - `refine`, `improve`, `deepen`, `expand`, `iterate`, `update` → Refine\n\n→ If unclear, ask the Purpose question above as first contextual question\n→ If clear, proceed to adaptive_analysis with inferred purpose\n\u003C\u002Fcritical_first_action>\n\n\u003Cadaptive_analysis>\nExtract and infer:\n\n- **Purpose**: Do, Plan, Research, or Refine\n- **Topic identifier**: Kebab-case identifier for file naming (e.g., `auth`, `stripe-payments`)\n- **Complexity**: Simple vs complex (affects prompt depth)\n- **Prompt structure**: Single vs multiple prompts\n- **Target** (Refine only): Which existing output to improve\n\nIf topic identifier not obvious, ask:\n- header: \"Topic\"\n- question: \"What topic\u002Ffeature is this for? (used for file naming)\"\n- Let user provide via \"Other\" option\n- Enforce kebab-case (convert spaces\u002Funderscores to hyphens)\n\nFor Refine purpose, also identify target output from `.prompts\u002F*\u002F` to improve.\n\u003C\u002Fadaptive_analysis>\n\n\u003Cchain_detection>\nScan `.prompts\u002F*\u002F` for existing `*-research.md` and `*-plan.md` files.\n\nIf found:\n1. List them: \"Found existing files: auth-research.md (in 001-auth-research\u002F), stripe-plan.md (in 005-stripe-plan\u002F)\"\n2. Use AskUserQuestion:\n   - header: \"Reference\"\n   - question: \"Should this prompt reference any existing research or plans?\"\n   - options: List found files + \"None\"\n   - multiSelect: true\n\nMatch by topic keyword when possible (e.g., \"auth plan\" → suggest auth-research.md).\n\u003C\u002Fchain_detection>\n\n\u003Ccontextual_questioning>\nGenerate 2-4 questions using AskUserQuestion based on purpose and gaps.\n\nLoad questions from: [references\u002Fquestion-bank.md](references\u002Fquestion-bank.md)\n\nRoute by purpose:\n- Do → artifact type, scope, approach\n- Plan → plan purpose, format, constraints\n- Research → depth, sources, output format\n- Refine → target selection, feedback, preservation\n\u003C\u002Fcontextual_questioning>\n\n\u003Cdecision_gate>\nAfter receiving answers, present decision gate using AskUserQuestion:\n\n- header: \"Ready\"\n- question: \"Ready to create the prompt?\"\n- options:\n  - \"Proceed\" - Create the prompt with current context\n  - \"Ask more questions\" - I have more details to clarify\n  - \"Let me add context\" - I want to provide additional information\n\nLoop until \"Proceed\" selected.\n\u003C\u002Fdecision_gate>\n\n\u003Cfinalization>\nAfter \"Proceed\" selected, state confirmation:\n\n\"Creating a {purpose} prompt for: {topic}\nFolder: .prompts\u002F{number}-{topic}-{purpose}\u002F\nReferences: {list any chained files}\"\n\nThen proceed to generation.\n\u003C\u002Ffinalization>\n\u003C\u002Fstep_0_intake_gate>\n\n\u003Cstep_1_generate>\n\u003Ctitle>Generate Prompt\u003C\u002Ftitle>\n\nLoad purpose-specific patterns:\n- Do: [references\u002Fdo-patterns.md](references\u002Fdo-patterns.md)\n- Plan: [references\u002Fplan-patterns.md](references\u002Fplan-patterns.md)\n- Research: [references\u002Fresearch-patterns.md](references\u002Fresearch-patterns.md)\n- Refine: [references\u002Frefine-patterns.md](references\u002Frefine-patterns.md)\n\nLoad intelligence rules: [references\u002Fintelligence-rules.md](references\u002Fintelligence-rules.md)\n\n\u003Cprompt_structure>\nAll generated prompts include:\n\n1. **Objective**: What to accomplish, why it matters\n2. **Context**: Referenced files (@), dynamic context (!)\n3. **Requirements**: Specific instructions for the task\n4. **Output specification**: Where to save, what structure\n5. **Metadata requirements**: For research\u002Fplan outputs, specify XML metadata structure\n6. **SUMMARY.md requirement**: All prompts must create a SUMMARY.md file\n7. **Success criteria**: How to know it worked\n\nFor Research and Plan prompts, output must include:\n- `\u003Cconfidence>` - How confident in findings\n- `\u003Cdependencies>` - What's needed to proceed\n- `\u003Copen_questions>` - What remains uncertain\n- `\u003Cassumptions>` - What was assumed\n\nAll prompts must create `SUMMARY.md` with:\n- **One-liner** - Substantive description of outcome\n- **Version** - v1 or iteration info\n- **Key Findings** - Actionable takeaways\n- **Files Created** - (Do prompts only)\n- **Decisions Needed** - What requires user input\n- **Blockers** - External impediments\n- **Next Step** - Concrete forward action\n\u003C\u002Fprompt_structure>\n\n\u003Cfile_creation>\n1. Create folder: `.prompts\u002F{number}-{topic}-{purpose}\u002F`\n2. Create `completed\u002F` subfolder\n3. Write prompt to: `.prompts\u002F{number}-{topic}-{purpose}\u002F{number}-{topic}-{purpose}.md`\n4. Prompt instructs output to: `.prompts\u002F{number}-{topic}-{purpose}\u002F{topic}-{purpose}.md`\n\u003C\u002Ffile_creation>\n\u003C\u002Fstep_1_generate>\n\n\u003Cstep_2_present>\n\u003Ctitle>Present Decision Tree\u003C\u002Ftitle>\n\nAfter saving prompt(s), present inline (not AskUserQuestion):\n\n\u003Csingle_prompt_presentation>\n```\nPrompt created: .prompts\u002F{number}-{topic}-{purpose}\u002F{number}-{topic}-{purpose}.md\n\nWhat's next?\n\n1. Run prompt now\n2. Review\u002Fedit prompt first\n3. Save for later\n4. Other\n\nChoose (1-4): _\n```\n\u003C\u002Fsingle_prompt_presentation>\n\n\u003Cmulti_prompt_presentation>\n```\nPrompts created:\n- .prompts\u002F001-auth-research\u002F001-auth-research.md\n- .prompts\u002F002-auth-plan\u002F002-auth-plan.md\n- .prompts\u002F003-auth-implement\u002F003-auth-implement.md\n\nDetected execution order: Sequential (002 references 001 output, 003 references 002 output)\n\nWhat's next?\n\n1. Run all prompts (sequential)\n2. Review\u002Fedit prompts first\n3. Save for later\n4. Other\n\nChoose (1-4): _\n```\n\u003C\u002Fmulti_prompt_presentation>\n\u003C\u002Fstep_2_present>\n\n\u003Cstep_3_execute>\n\u003Ctitle>Execution Engine\u003C\u002Ftitle>\n\n\u003Cexecution_modes>\n\u003Csingle_prompt>\nStraightforward execution of one prompt.\n\n1. Read prompt file contents\n2. Spawn Task agent with subagent_type=\"general-purpose\"\n3. Include in task prompt:\n   - The complete prompt contents\n   - Output location: `.prompts\u002F{number}-{topic}-{purpose}\u002F{topic}-{purpose}.md`\n4. Wait for completion\n5. Validate output (see validation section)\n6. Archive prompt to `completed\u002F` subfolder\n7. Report results with next-step options\n\u003C\u002Fsingle_prompt>\n\n\u003Csequential_execution>\nFor chained prompts where each depends on previous output.\n\n1. Build execution queue from dependency order\n2. For each prompt in queue:\n   a. Read prompt file\n   b. Spawn Task agent\n   c. Wait for completion\n   d. Validate output\n   e. If validation fails → stop, report failure, offer recovery options\n   f. If success → archive prompt, continue to next\n3. Report consolidated results\n\n\u003Cprogress_reporting>\nShow progress during execution:\n```\nExecuting 1\u002F3: 001-auth-research... ✓\nExecuting 2\u002F3: 002-auth-plan... ✓\nExecuting 3\u002F3: 003-auth-implement... (running)\n```\n\u003C\u002Fprogress_reporting>\n\u003C\u002Fsequential_execution>\n\n\u003Cparallel_execution>\nFor independent prompts with no dependencies.\n\n1. Read all prompt files\n2. **CRITICAL**: Spawn ALL Task agents in a SINGLE message\n   - This is required for true parallel execution\n   - Each task includes its output location\n3. Wait for all to complete\n4. Validate all outputs\n5. Archive all prompts\n6. Report consolidated results (successes and failures)\n\n\u003Cfailure_handling>\nUnlike sequential, parallel continues even if some fail:\n- Collect all results\n- Archive successful prompts\n- Report failures with details\n- Offer to retry failed prompts\n\u003C\u002Ffailure_handling>\n\u003C\u002Fparallel_execution>\n\n\u003Cmixed_dependencies>\nFor complex DAGs (e.g., two parallel research → one plan).\n\n1. Analyze dependency graph from @ references\n2. Group into execution layers:\n   - Layer 1: No dependencies (run parallel)\n   - Layer 2: Depends only on layer 1 (run after layer 1 completes)\n   - Layer 3: Depends on layer 2, etc.\n3. Execute each layer:\n   - Parallel within layer\n   - Sequential between layers\n4. Stop if any dependency fails (downstream prompts can't run)\n\n\u003Cexample>\n```\nLayer 1 (parallel): 001-api-research, 002-db-research\nLayer 2 (after layer 1): 003-architecture-plan\nLayer 3 (after layer 2): 004-implement\n```\n\u003C\u002Fexample>\n\u003C\u002Fmixed_dependencies>\n\u003C\u002Fexecution_modes>\n\n\u003Cdependency_detection>\n\u003Cautomatic_detection>\nScan prompt contents for @ references to determine dependencies:\n\n1. Parse each prompt for `@.prompts\u002F{number}-{topic}\u002F` patterns\n2. Build dependency graph\n3. Detect cycles (error if found)\n4. Determine execution order\n\n\u003Cinference_rules>\nIf no explicit @ references found, infer from purpose:\n- Research prompts: No dependencies (can parallel)\n- Plan prompts: Depend on same-topic research\n- Do prompts: Depend on same-topic plan\n\nOverride with explicit references when present.\n\u003C\u002Finference_rules>\n\u003C\u002Fautomatic_detection>\n\n\u003Cmissing_dependencies>\nIf a prompt references output that doesn't exist:\n\n1. Check if it's another prompt in this session (will be created)\n2. Check if it exists in `.prompts\u002F*\u002F` (already completed)\n3. If truly missing:\n   - Warn user: \"002-auth-plan references auth-research.md which doesn't exist\"\n   - Offer: Create the missing research prompt first? \u002F Continue anyway? \u002F Cancel?\n\u003C\u002Fmissing_dependencies>\n\u003C\u002Fdependency_detection>\n\n\u003Cvalidation>\n\u003Coutput_validation>\nAfter each prompt completes, verify success:\n\n1. **File exists**: Check output file was created\n2. **Not empty**: File has content (> 100 chars)\n3. **Metadata present** (for research\u002Fplan): Check for required XML tags\n   - `\u003Cconfidence>`\n   - `\u003Cdependencies>`\n   - `\u003Copen_questions>`\n   - `\u003Cassumptions>`\n4. **SUMMARY.md exists**: Check SUMMARY.md was created\n5. **SUMMARY.md complete**: Has required sections (Key Findings, Decisions Needed, Blockers, Next Step)\n6. **One-liner is substantive**: Not generic like \"Research completed\"\n\n\u003Cvalidation_failure>\nIf validation fails:\n- Report what's missing\n- Offer options:\n  - Retry the prompt\n  - Continue anyway (for non-critical issues)\n  - Stop and investigate\n\u003C\u002Fvalidation_failure>\n\u003C\u002Foutput_validation>\n\u003C\u002Fvalidation>\n\n\u003Cfailure_handling>\n\u003Csequential_failure>\nStop the chain immediately:\n```\n✗ Failed at 2\u002F3: 002-auth-plan\n\nCompleted:\n- 001-auth-research ✓ (archived)\n\nFailed:\n- 002-auth-plan: Output file not created\n\nNot started:\n- 003-auth-implement\n\nWhat's next?\n1. Retry 002-auth-plan\n2. View error details\n3. Stop here (keep completed work)\n4. Other\n```\n\u003C\u002Fsequential_failure>\n\n\u003Cparallel_failure>\nContinue others, report all results:\n```\nParallel execution completed with errors:\n\n✓ 001-api-research (archived)\n✗ 002-db-research: Validation failed - missing \u003Cconfidence> tag\n✓ 003-ui-research (archived)\n\nWhat's next?\n1. Retry failed prompt (002)\n2. View error details\n3. Continue without 002\n4. Other\n```\n\u003C\u002Fparallel_failure>\n\u003C\u002Ffailure_handling>\n\n\u003Carchiving>\n\u003Carchive_timing>\n- **Sequential**: Archive each prompt immediately after successful completion\n  - Provides clear state if execution stops mid-chain\n- **Parallel**: Archive all at end after collecting results\n  - Keeps prompts available for potential retry\n\n\u003Carchive_operation>\nMove prompt file to completed subfolder:\n```bash\nmv .prompts\u002F{number}-{topic}-{purpose}\u002F{number}-{topic}-{purpose}.md \\\n   .prompts\u002F{number}-{topic}-{purpose}\u002Fcompleted\u002F\n```\n\nOutput file stays in place (not moved).\n\u003C\u002Farchive_operation>\n\u003C\u002Farchiving>\n\n\u003Cresult_presentation>\n\u003Csingle_result>\n```\n✓ Executed: 001-auth-research\n✓ Created: .prompts\u002F001-auth-research\u002FSUMMARY.md\n\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n# Auth Research Summary\n\n**JWT with jose library and httpOnly cookies recommended**\n\n## Key Findings\n• jose outperforms jsonwebtoken with better TypeScript support\n• httpOnly cookies required (localStorage is XSS vulnerable)\n• Refresh rotation is OWASP standard\n\n## Decisions Needed\nNone - ready for planning\n\n## Blockers\nNone\n\n## Next Step\nCreate auth-plan.md\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\nWhat's next?\n1. Create planning prompt (auth-plan)\n2. View full research output\n3. Done\n4. Other\n```\n\nDisplay the actual SUMMARY.md content inline so user sees findings without opening files.\n\u003C\u002Fsingle_result>\n\n\u003Cchain_result>\n```\n✓ Chain completed: auth workflow\n\nResults:\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n001-auth-research\n**JWT with jose library and httpOnly cookies recommended**\nDecisions: None • Blockers: None\n\n002-auth-plan\n**4-phase implementation: types → JWT core → refresh → tests**\nDecisions: Approve 15-min token expiry • Blockers: None\n\n003-auth-implement\n**JWT middleware complete with 6 files created**\nDecisions: Review before Phase 2 • Blockers: None\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\nAll prompts archived. Full summaries in .prompts\u002F*\u002FSUMMARY.md\n\nWhat's next?\n1. Review implementation\n2. Run tests\n3. Create new prompt chain\n4. Other\n```\n\nFor chains, show condensed one-liner from each SUMMARY.md with decisions\u002Fblockers flagged.\n\u003C\u002Fchain_result>\n\u003C\u002Fresult_presentation>\n\n\u003Cspecial_cases>\n\u003Cre_running_completed>\nIf user wants to re-run an already-completed prompt:\n\n1. Check if prompt is in `completed\u002F` subfolder\n2. Move it back to parent folder\n3. Optionally backup existing output: `{output}.bak`\n4. Execute normally\n\u003C\u002Fre_running_completed>\n\n\u003Coutput_conflicts>\nIf output file already exists:\n\n1. For re-runs: Backup existing → `{filename}.bak`\n2. For new runs: Should not happen (unique numbering)\n3. If conflict detected: Ask user - Overwrite? \u002F Rename? \u002F Cancel?\n\u003C\u002Foutput_conflicts>\n\n\u003Ccommit_handling>\nAfter successful execution:\n\n1. Do NOT auto-commit (user controls git workflow)\n2. Mention what files were created\u002Fmodified\n3. User can commit when ready\n\nException: If user explicitly requests commit, stage and commit:\n- Output files created\n- Prompts archived\n- Any implementation changes (for Do prompts)\n\u003C\u002Fcommit_handling>\n\n\u003Crecursive_prompts>\nIf a prompt's output includes instructions to create more prompts:\n\n1. This is advanced usage - don't auto-detect\n2. Present the output to user\n3. User can invoke skill again to create follow-up prompts\n4. Maintains user control over prompt creation\n\u003C\u002Frecursive_prompts>\n\u003C\u002Fspecial_cases>\n\u003C\u002Fstep_3_execute>\n\n\u003C\u002Fautomated_workflow>\n\n\u003Creference_guides>\n**Prompt patterns by purpose:**\n- [references\u002Fdo-patterns.md](references\u002Fdo-patterns.md) - Execution prompts + output structure\n- [references\u002Fplan-patterns.md](references\u002Fplan-patterns.md) - Planning prompts + plan.md structure\n- [references\u002Fresearch-patterns.md](references\u002Fresearch-patterns.md) - Research prompts + research.md structure\n- [references\u002Frefine-patterns.md](references\u002Frefine-patterns.md) - Iteration prompts + versioning\n\n**Shared templates:**\n- [references\u002Fsummary-template.md](references\u002Fsummary-template.md) - SUMMARY.md structure and field requirements\n- [references\u002Fmetadata-guidelines.md](references\u002Fmetadata-guidelines.md) - Confidence, dependencies, open questions, assumptions\n\n**Supporting references:**\n- [references\u002Fquestion-bank.md](references\u002Fquestion-bank.md) - Intake questions by purpose\n- [references\u002Fintelligence-rules.md](references\u002Fintelligence-rules.md) - Extended thinking, parallel tools, depth decisions\n\u003C\u002Freference_guides>\n\n\u003Csuccess_criteria>\n**Prompt Creation:**\n- Intake gate completed with purpose and topic identified\n- Chain detection performed, relevant files referenced\n- Prompt generated with correct structure for purpose\n- Folder created in `.prompts\u002F` with correct naming\n- Output file location specified in prompt\n- SUMMARY.md requirement included in prompt\n- Metadata requirements included for Research\u002FPlan outputs\n- Quality controls included for Research outputs (verification checklist, QA, pre-submission)\n- Streaming write instructions included for Research outputs\n- Decision tree presented\n\n**Execution (if user chooses to run):**\n- Dependencies correctly detected and ordered\n- Prompts executed in correct order (sequential\u002Fparallel\u002Fmixed)\n- Output validated after each completion\n- SUMMARY.md created with all required sections\n- One-liner is substantive (not generic)\n- Failed prompts handled gracefully with recovery options\n- Successful prompts archived to `completed\u002F` subfolder\n- SUMMARY.md displayed inline in results\n- Results presented with decisions\u002Fblockers flagged\n\n**Research Quality (for Research prompts):**\n- Verification checklist completed\n- Quality report distinguishes verified from assumed claims\n- Sources consulted listed with URLs\n- Confidence levels assigned to findings\n- Critical claims verified with official documentation\n\u003C\u002Fsuccess_criteria>\n",{"name":5,"origin":24,"description":7},"https:\u002F\u002Fraw.githubusercontent.com\u002Fglittercowboy\u002Ftaches-cc-resources\u002Fmain\u002Fskills\u002Fcreate-meta-prompts\u002FSKILL.md","main","481dcd7dd2e288b9d627a2bd017db70e64e6c6fa","core\u002Fskills\u002Ftaches--create-meta-prompts",2,"https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI","https:\u002F\u002Fyana.vutam.link\u002F","https:\u002F\u002Fgithub.com\u002FHoronomos-laps","https:\u002F\u002Fyana.vutam.link","An open-source governed runtime & execution environment for multi-agent AI systems.","2026-11-12 23:12:27",{"source":8,"skillName":5,"commands":36,"prompt":46,"zipAvailable":47},[37,40,43],{"tool":38,"command":39},"npx","npx skills add Horonomos-laps\u002FYana-AI --skill create-meta-prompts",{"tool":41,"command":42},"bunx","bunx skills add Horonomos-laps\u002FYana-AI --skill create-meta-prompts",{"tool":44,"command":45},"pnpm","pnpm dlx skills add Horonomos-laps\u002FYana-AI --skill create-meta-prompts","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002Fcore\u002Fskills\u002Ftaches--create-meta-prompts\u002FSKILL.md",true,[],[50,58,65,73,81,89],{"skillNo":51,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":54,"licenseFileUrl":6,"sourceHtmlUrl":55,"categories":56,"keywords":57,"publishedAt":19},"01M735GKET5YVYKAY62W2P5YXB","openai--sentry--sentry","Use when the user asks to inspect Sentry issues or events, summarize recent production errors, or pull basic Sentry health data via the Sentry API; perform read-only queries with the bundled script and require `SENTRY_AUTH_TOKEN`.","MIT","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fopenai--sentry--sentry\u002FSKILL.md",[],[],{"skillNo":59,"name":60,"nameZh":6,"description":61,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":62,"categories":63,"keywords":64,"publishedAt":19},"01M73DCAP83T17HY7SMGK7MGFE","ito-basket-compare","Compare Itô prediction-market baskets against a user's knowledge base, portfolio notes, financial context, watchlist, or research thesis. Use for read-only basket comparison and gap analysis without investment advice or live trading.","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002Fcore\u002Fskills\u002Fito-basket-compare\u002FSKILL.md",[],[],{"skillNo":66,"name":67,"nameZh":6,"description":68,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":69,"sourceHtmlUrl":70,"categories":71,"keywords":72,"publishedAt":19},"01M734CYEMZFDR3KH6SW0JP8TD","implementing-network-traffic-baselining","Build network traffic baselines from NetFlow\u002FIPFIX data using Python pandas for statistical analysis, z-score anomaly detection, and hourly\u002Fdaily traffic pattern profiling","https:\u002F\u002Fgithub.com\u002Fyanacuti1121\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fimplementing-network-traffic-baselining\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fimplementing-network-traffic-baselining\u002FSKILL.md",[],[],{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":77,"sourceHtmlUrl":78,"categories":79,"keywords":80,"publishedAt":19},"01M73ASZ6NJRJM44Z2P84NGQ3Q","detecting-evasion-techniques-in-endpoint-logs","Detects defense evasion techniques used by adversaries in endpoint logs including log tampering, timestomping, process injection, and security tool disabling. Use when investigating suspicious endpoint behavior, building detection rules for evasion tactics, or conducting threat hunting for stealthy adversary activity. Activates for requests involving evasion detection, defense evasion analysis, log tampering detection, or MITRE ATT&CK TA0005.","https:\u002F\u002Fgithub.com\u002Fyanacuti1121\u002FYana-AI\u002Fblob\u002Fmain\u002Fcore\u002Fskills\u002Fdetecting-evasion-techniques-in-endpoint-logs\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002Fcore\u002Fskills\u002Fdetecting-evasion-techniques-in-endpoint-logs\u002FSKILL.md",[],[],{"skillNo":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":85,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M734AFNXD06Z78QRK0VKGT7Q","implementing-microsegmentation-with-guardicore","Implementing microsegmentation using Akamai Guardicore Segmentation to map application dependencies, create granular network policies, visualize east-west traffic flows, and enforce least-privilege communication between workloads across data centers and cloud.","https:\u002F\u002Fgithub.com\u002Fyanacuti1121\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fimplementing-microsegmentation-with-guardicore\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fimplementing-microsegmentation-with-guardicore\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":92,"categories":93,"keywords":94,"publishedAt":19},"01M736TK27QMCXK0DVCH25SJN5","ramdisk-tmpfs-patterns","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Framdisk-tmpfs-patterns\u002FSKILL.md",[],[]]