[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M72ZCB9SKR5NGQHDNS2AC91Q":3,"skill-same:01M72ZCB9SKR5NGQHDNS2AC91Q":76},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":26,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":43},"01M72ZCB9SKR5NGQHDNS2AC91Q","bmad-distillator",null,"Lossless LLM-optimized compression of source documents. Use when the user requests to 'distill documents' or 'create a distillate'.","developAgent\u002Fbmad-start","developAgent","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F264530239?v=4","01M71J8X6CY24V5ZZHFGETRMSW","01M71J8X739FS28DGQAAY7V4QR",1,"MIT","https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002F.windsurf\u002Fskills\u002Fbmad-distillator\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: bmad-distillator\ndescription: Lossless LLM-optimized compression of source documents. Use when the user requests to 'distill documents' or 'create a distillate'.\n---\n\n# Distillator: A Document Distillation Engine\n\n## Overview\n\nThis skill produces hyper-compressed, token-efficient documents (distillates) from any set of source documents. A distillate preserves every fact, decision, constraint, and relationship from the sources while stripping all overhead that humans need and LLMs don't. Act as an information extraction and compression specialist. The output is a single dense document (or semantically-split set) that a downstream LLM workflow can consume as sole context input without information loss.\n\nThis is a compression task, not a summarization task. Summaries are lossy. Distillates are lossless compression optimized for LLM consumption.\n\n## On Activation\n\n1. **Validate inputs.** The caller must provide:\n   - **source_documents** (required) — One or more file paths, folder paths, or glob patterns to distill\n   - **downstream_consumer** (optional) — What workflow\u002Fagent consumes this distillate (e.g., \"PRD creation\", \"architecture design\"). When provided, use it to judge signal vs noise. When omitted, preserve everything.\n   - **token_budget** (optional) — Approximate target size. When provided and the distillate would exceed it, trigger semantic splitting.\n   - **output_path** (optional) — Where to save. When omitted, save adjacent to the primary source document with `-distillate.md` suffix.\n   - **--validate** (flag) — Run round-trip reconstruction test after producing the distillate.\n\n2. **Route** — proceed to Stage 1.\n\n## Stages\n\n| # | Stage | Purpose |\n|---|-------|---------|\n| 1 | Analyze | Run analysis script, determine routing and splitting |\n| 2 | Compress | Spawn compressor agent(s) to produce the distillate |\n| 3 | Verify & Output | Completeness check, format check, save output |\n| 4 | Round-Trip Validate | (--validate only) Reconstruct and diff against originals |\n\n### Stage 1: Analyze\n\nRun `scripts\u002Fanalyze_sources.py --help` then run it with the source paths. Use its routing recommendation and grouping output to drive Stage 2. Do NOT read the source documents yourself.\n\n### Stage 2: Compress\n\n**Single mode** (routing = `\"single\"`, ≤3 files, ≤15K estimated tokens):\n\nSpawn one subagent using `agents\u002Fdistillate-compressor.md` with all source file paths.\n\n**Fan-out mode** (routing = `\"fan-out\"`):\n\n1. Spawn one compressor subagent per group from the analysis output. Each compressor receives only its group's file paths and produces an intermediate distillate.\n\n2. After all compressors return, spawn one final **merge compressor** subagent using `agents\u002Fdistillate-compressor.md`. Pass it the intermediate distillate contents as its input (not the original files). Its job is cross-group deduplication, thematic regrouping, and final compression.\n\n3. Clean up intermediate distillate content (it exists only in memory, not saved to disk).\n\n**Graceful degradation:** If subagent spawning is unavailable, read the source documents and perform the compression work directly using the same instructions from `agents\u002Fdistillate-compressor.md`. For fan-out, process groups sequentially then merge.\n\nThe compressor returns a structured JSON result containing the distillate content, source headings, named entities, and token estimate.\n\n### Stage 3: Verify & Output\n\nAfter the compressor (or merge compressor) returns:\n\n1. **Completeness check.** Using the headings and named entities list returned by the compressor, verify each appears in the distillate content. If gaps are found, send them back to the compressor for a targeted fix pass — not a full recompression. Limit to 2 fix passes maximum.\n\n2. **Format check.** Verify the output follows distillate format rules:\n   - No prose paragraphs (only bullets)\n   - No decorative formatting\n   - No repeated information\n   - Each bullet is self-contained\n   - Themes are clearly delineated with `##` headings\n\n3. **Determine output format.** Using the split prediction from Stage 1 and actual distillate size:\n\n   **Single distillate** (≤~5,000 tokens or token_budget not exceeded):\n\n   Save as a single file with frontmatter:\n\n   ```yaml\n   ---\n   type: bmad-distillate\n   sources:\n     - \"{relative path to source file 1}\"\n     - \"{relative path to source file 2}\"\n   downstream_consumer: \"{consumer or 'general'}\"\n   created: \"{date}\"\n   token_estimate: {approximate token count}\n   parts: 1\n   ---\n   ```\n\n   **Split distillate** (>~5,000 tokens, or token_budget requires it):\n\n   Create a folder `{base-name}-distillate\u002F` containing:\n\n   ```\n   {base-name}-distillate\u002F\n   ├── _index.md           # Orientation, cross-cutting items, section manifest\n   ├── 01-{topic-slug}.md  # Self-contained section\n   ├── 02-{topic-slug}.md\n   └── 03-{topic-slug}.md\n   ```\n\n   The `_index.md` contains:\n   - Frontmatter with sources (relative paths from the distillate folder to the originals)\n   - 3-5 bullet orientation (what was distilled, from what)\n   - Section manifest: each section's filename + 1-line description\n   - Cross-cutting items that span multiple sections\n\n   Each section file is self-contained — loadable independently. Include a 1-line context header: \"This section covers [topic]. Part N of M.\"\n\n   Source paths in frontmatter must be relative to the distillate's location.\n\n4. **Measure distillate.** Run `scripts\u002Fanalyze_sources.py` on the final distillate file(s) to get accurate token counts for the output. Use the `total_estimated_tokens` from this analysis as `distillate_total_tokens`.\n\n5. **Report results.** Always return structured JSON output:\n\n   ```json\n   {\n     \"status\": \"complete\",\n     \"distillate\": \"{path or folder path}\",\n     \"section_distillates\": [\"{path1}\", \"{path2}\"] or null,\n     \"source_total_tokens\": N,\n     \"distillate_total_tokens\": N,\n     \"compression_ratio\": \"X:1\",\n     \"source_documents\": [\"{path1}\", \"{path2}\"],\n     \"completeness_check\": \"pass\" or \"pass_with_additions\"\n   }\n   ```\n\n   Where `source_total_tokens` is from the Stage 1 analysis and `distillate_total_tokens` is from step 4. The `compression_ratio` is `source_total_tokens \u002F distillate_total_tokens` formatted as \"X:1\" (e.g., \"3.2:1\").\n\n6. If `--validate` flag was set, proceed to Stage 4. Otherwise, done.\n\n### Stage 4: Round-Trip Validation (--validate only)\n\nThis stage proves the distillate is lossless by reconstructing source documents from the distillate alone. Use for critical documents where information loss is unacceptable, or as a quality gate for high-stakes downstream workflows. Not for routine use — it adds significant token cost.\n\n1. **Spawn the reconstructor agent** using `agents\u002Fround-trip-reconstructor.md`. Pass it ONLY the distillate file path (or `_index.md` path for split distillates) — it must NOT have access to the original source documents.\n\n   For split distillates, spawn one reconstructor per section in parallel. Each receives its section file plus the `_index.md` for cross-cutting context.\n\n   **Graceful degradation:** If subagent spawning is unavailable, this stage cannot be performed by the main agent (it has already seen the originals). Report that round-trip validation requires subagent support and skip.\n\n2. **Receive reconstructions.** The reconstructor returns reconstruction file paths saved adjacent to the distillate.\n\n3. **Perform semantic diff.** Read both the original source documents and the reconstructions. For each section of the original, assess:\n   - Is the core information present in the reconstruction?\n   - Are specific details preserved (numbers, names, decisions)?\n   - Are relationships and rationale intact?\n   - Did the reconstruction add anything not in the original? (indicates hallucination filling gaps)\n\n4. **Produce validation report** saved adjacent to the distillate as `-validation-report.md`:\n\n   ```markdown\n   ---\n   type: distillate-validation\n   distillate: \"{distillate path}\"\n   sources: [\"{source paths}\"]\n   created: \"{date}\"\n   ---\n\n   ## Validation Summary\n   - Status: PASS | PASS_WITH_WARNINGS | FAIL\n   - Information preserved: {percentage estimate}\n   - Gaps found: {count}\n   - Hallucinations detected: {count}\n\n   ## Gaps (information in originals but missing from reconstruction)\n   - {gap description} — Source: {which original}, Section: {where}\n\n   ## Hallucinations (information in reconstruction not traceable to originals)\n   - {hallucination description} — appears to fill gap in: {section}\n\n   ## Possible Gap Markers (flagged by reconstructor)\n   - {marker description}\n   ```\n\n5. **If gaps are found**, offer to run a targeted fix pass on the distillate — adding the missing information without full recompression. Limit to 2 fix passes maximum.\n\n6. **Clean up** — delete the temporary reconstruction files after the report is generated.","# Distillator: A Document Distillation Engine\n\n## Overview\n\nThis skill produces hyper-compressed, token-efficient documents (distillates) from any set of source documents. A distillate preserves every fact, decision, constraint, and relationship from the sources while stripping all overhead that humans need and LLMs don't. Act as an information extraction and compression specialist. The output is a single dense document (or semantically-split set) that a downstream LLM workflow can consume as sole context input without information loss.\n\nThis is a compression task, not a summarization task. Summaries are lossy. Distillates are lossless compression optimized for LLM consumption.\n\n## On Activation\n\n1. **Validate inputs.** The caller must provide:\n   - **source_documents** (required) — One or more file paths, folder paths, or glob patterns to distill\n   - **downstream_consumer** (optional) — What workflow\u002Fagent consumes this distillate (e.g., \"PRD creation\", \"architecture design\"). When provided, use it to judge signal vs noise. When omitted, preserve everything.\n   - **token_budget** (optional) — Approximate target size. When provided and the distillate would exceed it, trigger semantic splitting.\n   - **output_path** (optional) — Where to save. When omitted, save adjacent to the primary source document with `-distillate.md` suffix.\n   - **--validate** (flag) — Run round-trip reconstruction test after producing the distillate.\n\n2. **Route** — proceed to Stage 1.\n\n## Stages\n\n| # | Stage | Purpose |\n|---|-------|---------|\n| 1 | Analyze | Run analysis script, determine routing and splitting |\n| 2 | Compress | Spawn compressor agent(s) to produce the distillate |\n| 3 | Verify & Output | Completeness check, format check, save output |\n| 4 | Round-Trip Validate | (--validate only) Reconstruct and diff against originals |\n\n### Stage 1: Analyze\n\nRun `scripts\u002Fanalyze_sources.py --help` then run it with the source paths. Use its routing recommendation and grouping output to drive Stage 2. Do NOT read the source documents yourself.\n\n### Stage 2: Compress\n\n**Single mode** (routing = `\"single\"`, ≤3 files, ≤15K estimated tokens):\n\nSpawn one subagent using `agents\u002Fdistillate-compressor.md` with all source file paths.\n\n**Fan-out mode** (routing = `\"fan-out\"`):\n\n1. Spawn one compressor subagent per group from the analysis output. Each compressor receives only its group's file paths and produces an intermediate distillate.\n\n2. After all compressors return, spawn one final **merge compressor** subagent using `agents\u002Fdistillate-compressor.md`. Pass it the intermediate distillate contents as its input (not the original files). Its job is cross-group deduplication, thematic regrouping, and final compression.\n\n3. Clean up intermediate distillate content (it exists only in memory, not saved to disk).\n\n**Graceful degradation:** If subagent spawning is unavailable, read the source documents and perform the compression work directly using the same instructions from `agents\u002Fdistillate-compressor.md`. For fan-out, process groups sequentially then merge.\n\nThe compressor returns a structured JSON result containing the distillate content, source headings, named entities, and token estimate.\n\n### Stage 3: Verify & Output\n\nAfter the compressor (or merge compressor) returns:\n\n1. **Completeness check.** Using the headings and named entities list returned by the compressor, verify each appears in the distillate content. If gaps are found, send them back to the compressor for a targeted fix pass — not a full recompression. Limit to 2 fix passes maximum.\n\n2. **Format check.** Verify the output follows distillate format rules:\n   - No prose paragraphs (only bullets)\n   - No decorative formatting\n   - No repeated information\n   - Each bullet is self-contained\n   - Themes are clearly delineated with `##` headings\n\n3. **Determine output format.** Using the split prediction from Stage 1 and actual distillate size:\n\n   **Single distillate** (≤~5,000 tokens or token_budget not exceeded):\n\n   Save as a single file with frontmatter:\n\n   ```yaml\n   ---\n   type: bmad-distillate\n   sources:\n     - \"{relative path to source file 1}\"\n     - \"{relative path to source file 2}\"\n   downstream_consumer: \"{consumer or 'general'}\"\n   created: \"{date}\"\n   token_estimate: {approximate token count}\n   parts: 1\n   ---\n   ```\n\n   **Split distillate** (>~5,000 tokens, or token_budget requires it):\n\n   Create a folder `{base-name}-distillate\u002F` containing:\n\n   ```\n   {base-name}-distillate\u002F\n   ├── _index.md           # Orientation, cross-cutting items, section manifest\n   ├── 01-{topic-slug}.md  # Self-contained section\n   ├── 02-{topic-slug}.md\n   └── 03-{topic-slug}.md\n   ```\n\n   The `_index.md` contains:\n   - Frontmatter with sources (relative paths from the distillate folder to the originals)\n   - 3-5 bullet orientation (what was distilled, from what)\n   - Section manifest: each section's filename + 1-line description\n   - Cross-cutting items that span multiple sections\n\n   Each section file is self-contained — loadable independently. Include a 1-line context header: \"This section covers [topic]. Part N of M.\"\n\n   Source paths in frontmatter must be relative to the distillate's location.\n\n4. **Measure distillate.** Run `scripts\u002Fanalyze_sources.py` on the final distillate file(s) to get accurate token counts for the output. Use the `total_estimated_tokens` from this analysis as `distillate_total_tokens`.\n\n5. **Report results.** Always return structured JSON output:\n\n   ```json\n   {\n     \"status\": \"complete\",\n     \"distillate\": \"{path or folder path}\",\n     \"section_distillates\": [\"{path1}\", \"{path2}\"] or null,\n     \"source_total_tokens\": N,\n     \"distillate_total_tokens\": N,\n     \"compression_ratio\": \"X:1\",\n     \"source_documents\": [\"{path1}\", \"{path2}\"],\n     \"completeness_check\": \"pass\" or \"pass_with_additions\"\n   }\n   ```\n\n   Where `source_total_tokens` is from the Stage 1 analysis and `distillate_total_tokens` is from step 4. The `compression_ratio` is `source_total_tokens \u002F distillate_total_tokens` formatted as \"X:1\" (e.g., \"3.2:1\").\n\n6. If `--validate` flag was set, proceed to Stage 4. Otherwise, done.\n\n### Stage 4: Round-Trip Validation (--validate only)\n\nThis stage proves the distillate is lossless by reconstructing source documents from the distillate alone. Use for critical documents where information loss is unacceptable, or as a quality gate for high-stakes downstream workflows. Not for routine use — it adds significant token cost.\n\n1. **Spawn the reconstructor agent** using `agents\u002Fround-trip-reconstructor.md`. Pass it ONLY the distillate file path (or `_index.md` path for split distillates) — it must NOT have access to the original source documents.\n\n   For split distillates, spawn one reconstructor per section in parallel. Each receives its section file plus the `_index.md` for cross-cutting context.\n\n   **Graceful degradation:** If subagent spawning is unavailable, this stage cannot be performed by the main agent (it has already seen the originals). Report that round-trip validation requires subagent support and skip.\n\n2. **Receive reconstructions.** The reconstructor returns reconstruction file paths saved adjacent to the distillate.\n\n3. **Perform semantic diff.** Read both the original source documents and the reconstructions. For each section of the original, assess:\n   - Is the core information present in the reconstruction?\n   - Are specific details preserved (numbers, names, decisions)?\n   - Are relationships and rationale intact?\n   - Did the reconstruction add anything not in the original? (indicates hallucination filling gaps)\n\n4. **Produce validation report** saved adjacent to the distillate as `-validation-report.md`:\n\n   ```markdown\n   ---\n   type: distillate-validation\n   distillate: \"{distillate path}\"\n   sources: [\"{source paths}\"]\n   created: \"{date}\"\n   ---\n\n   ## Validation Summary\n   - Status: PASS | PASS_WITH_WARNINGS | FAIL\n   - Information preserved: {percentage estimate}\n   - Gaps found: {count}\n   - Hallucinations detected: {count}\n\n   ## Gaps (information in originals but missing from reconstruction)\n   - {gap description} — Source: {which original}, Section: {where}\n\n   ## Hallucinations (information in reconstruction not traceable to originals)\n   - {hallucination description} — appears to fill gap in: {section}\n\n   ## Possible Gap Markers (flagged by reconstructor)\n   - {marker description}\n   ```\n\n5. **If gaps are found**, offer to run a targeted fix pass on the distillate — adding the missing information without full recompression. Limit to 2 fix passes maximum.\n\n6. **Clean up** — delete the temporary reconstruction files after the report is generated.",{"name":5,"description":7},"main","1907bbe850fb749efa8b499b8da7ebb78a663519",".windsurf\u002Fskills\u002Fbmad-distillator",0,"https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start","https:\u002F\u002Fgithub.com\u002FdevelopAgent","2026-07-19 05:19:22",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add developAgent\u002Fbmad-start --skill bmad-distillator",{"tool":36,"command":37},"bunx","bunx skills add developAgent\u002Fbmad-start --skill bmad-distillator",{"tool":39,"command":40},"pnpm","pnpm dlx skills add developAgent\u002Fbmad-start --skill bmad-distillator","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002F.windsurf\u002Fskills\u002Fbmad-distillator\u002FSKILL.md",true,[44,50,54,58,62,66,72],{"path":45,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":48,"sourceHtmlUrl":49},"agents\u002Fdistillate-compressor.md","REFERENCE","参考文档",4951,"https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002F.windsurf\u002Fskills\u002Fbmad-distillator\u002Fagents\u002Fdistillate-compressor.md",{"path":51,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":52,"sourceHtmlUrl":53},"agents\u002Fround-trip-reconstructor.md",2802,"https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002F.windsurf\u002Fskills\u002Fbmad-distillator\u002Fagents\u002Fround-trip-reconstructor.md",{"path":55,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":56,"sourceHtmlUrl":57},"resources\u002Fcompression-rules.md",2662,"https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002F.windsurf\u002Fskills\u002Fbmad-distillator\u002Fresources\u002Fcompression-rules.md",{"path":59,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":60,"sourceHtmlUrl":61},"resources\u002Fdistillate-format-reference.md",11880,"https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002F.windsurf\u002Fskills\u002Fbmad-distillator\u002Fresources\u002Fdistillate-format-reference.md",{"path":63,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":64,"sourceHtmlUrl":65},"resources\u002Fsplitting-strategy.md",3337,"https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002F.windsurf\u002Fskills\u002Fbmad-distillator\u002Fresources\u002Fsplitting-strategy.md",{"path":67,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":70,"sourceHtmlUrl":71},"scripts\u002Fanalyze_sources.py","SCRIPT","脚本",9912,"https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002F.windsurf\u002Fskills\u002Fbmad-distillator\u002Fscripts\u002Fanalyze_sources.py",{"path":73,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":74,"sourceHtmlUrl":75},"scripts\u002Ftests\u002Ftest_analyze_sources.py",7723,"https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002F.windsurf\u002Fskills\u002Fbmad-distillator\u002Fscripts\u002Ftests\u002Ftest_analyze_sources.py",[77,84,91,98,105,112],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":18},"01M72ZN1RJ28HE3TG8CK5VMDZ0","writing-plans","Use when you have a spec or requirements for a multi-step task, before touching code","https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002Fskills\u002Fwriting-plans\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":18},"01M72ZMZG4KQAFDEHG6KABQW8R","to-prd","Turn the current conversation into a PRD and publish it to the project issue tracker — no interview, just synthesis of what you've already discussed.","https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002Fskills\u002Fto-prd\u002FSKILL.md",[],[],{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":18},"01M72ZMWB9ZNA0Y5KGAHTW89E8","steve-jobs-perspective","史蒂夫·乔布斯(Steve Jobs)的思维框架与表达方式。基于Isaacson授权传记、Stanford演讲、\nLost Interview、D Conference系列、Make Something Wonderful、30+一手来源的深度调研，\n提炼6个核心心智模型、8条决策启发式和完整的表达DNA。\n用途：作为思维顾问，用乔布斯的视角分析产品、审视决策、提供反馈。\n当用户提到「用乔布斯的视角」「乔布斯会怎么看」「Jobs模式」「steve jobs perspective」时使用。\n即使用户只是说「帮我用乔布斯的角度想想」「如果乔布斯会怎么做」「切换到乔布斯」也应触发。","https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002Fskills\u002Fsteve-jobs-perspective\u002FSKILL.md",[],[],{"skillNo":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":18},"01M72ZMS3ZRQY9PCRPRVWY040M","pinme","Use this skill when the user mentions \"pinme\", or needs to upload files, store to IPFS, create\u002Fpublish\u002Fdeploy websites or full-stack services (including frontend pages, backend APIs, database storage, email sending, etc.), or any feature requiring backend database\u002Fserver support.","https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002Fskills\u002Fpinme\u002FSKILL.md",[],[],{"skillNo":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":109,"categories":110,"keywords":111,"publishedAt":18},"01M72ZMPJHQ8NHC9WSH41DKV1F","pinme-share","Use this skill when the user wants to share, publish, or upload a static result through PinMe, especially by generating a static HTML share page for a PinMe project link, deployed full-stack app, Codex conversation summary, report, file, demo, or any 分享\u002F发布\u002F上传分享页 request that should end with `pinme upload`.","https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002Fskills\u002Fpinme-share\u002FSKILL.md",[],[],{"skillNo":113,"name":114,"nameZh":6,"description":115,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":116,"categories":117,"keywords":118,"publishedAt":18},"01M72ZMM4QBVWMAW0YYM7H0Y6P","pinme-llm","Use this skill when a PinMe project (Worker TypeScript) needs to call OpenRouter-backed LLM APIs, including models, chat\u002Fcompletions, streaming, or OpenRouter web search. Guides AI to generate correct Worker TS code.","https:\u002F\u002Fgithub.com\u002FdevelopAgent\u002Fbmad-start\u002Fblob\u002Fmain\u002Fskills\u002Fpinme-llm\u002FSKILL.md",[],[]]