[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M74MWYTHT498SVMN823896T7":3,"skill-same:01M74MWYTHT498SVMN823896T7":53},{"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":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":6,"ownerDescription":31,"updatedAt":32,"install":33,"resources":46},"01M74MWYTHT498SVMN823896T7","workflow-skill-creator",null,"Distills a completed user workflow or interaction into a reusable agent skill. Use when the user asks to turn their workflow, interaction, or multi-step process into a skill, or when they say \"make this a skill\", \"create a skill from what we just did\", \"package this workflow\" or similar. Do not use for creating skills from scratch without an existing workflow (use a generic skill-creator for that).","pengqianhan\u002FAI-Human-Research-OS","pengqianhan","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F4286531?v=4","01M74M3DT9HNZ6HPAK8TX69EX9","PengqianHan","01M74M3DTVSN2GXW0JH2W396FK",3,"MIT","https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fscience-skills\u002Fworkflow_skill_creator\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: workflow-skill-creator\ndescription: >\n  Distills a completed user workflow or interaction into a reusable agent\n  skill. Use when the user asks to turn their workflow, interaction, or\n  multi-step process into a skill, or when they say \"make this a skill\",\n  \"create a skill from what we just did\", \"package this workflow\" or similar.\n  Do not use for creating skills from scratch without an existing workflow\n  (use a generic skill-creator for that).\n---\n\n# Workflow-to-Skill Distiller\n\nTurns a completed workflow into a reusable agent skill. Specifically, this skill\nextracts patterns from an interaction or workflow that **already happened** and\npackages them.\n\n> [!CAUTION] **You MUST complete Phase 1 (Brainstorming) before writing any code\n> or SKILL.md content.** Skipping brainstorming produces skills that are either\n> too rigid or too vague. The brainstorming conversation is the most important\n> part of this process.\n\n## Phase 1: Brainstorming (MANDATORY)\n\nHave an **iterative back-and-forth conversation** with the user. Do NOT ask all\nquestions at once. Pick 2-3 relevant questions per round from the bank below,\nrefine your understanding, and ask follow-ups.\n\n### Round 1: Understand the Workflow\n\nStart by summarizing what you observed from the workflow, then ask:\n\n1.  \"Here's my understanding of the workflow: [summary]. Is this accurate? What\n    would you change?\"\n2.  \"What are the expected inputs and outputs for this workflow?\"\n3.  \"How often do you expect to run this workflow? Is it recurring or one-off?\"\n\n### Round 2: Flexibility and Error Handling\n\nFor each step identified in the workflow, determine its rigidity:\n\n1.  \"For [step X], if the primary approach fails (e.g., API down, no results),\n    should the agent: (a) ask you for guidance, (b) try alternative approaches\n    automatically, or (c) fail loudly with an error?\"\n2.  \"Are there any steps where the exact method matters (e.g., must use a\n    specific database), vs. steps where any reasonable approach is fine?\"\n3.  \"Should the skill handle edge cases silently or surface them to the user?\"\n\n### Round 3: Dependencies and Resources\n\nBefore asking these questions, check which of your installed skills overlap with\nthe workflow. If an existing skill from the science bundle covers a step, the\nnew skill **MUST** reference it — do not offer a self-contained option.\n\n1.  \"I noticed the workflow uses functionality covered by [existing skill X,\n    skill Y]. The new skill will reference these rather than reimplementing\n    them. Are there any other tools or skills you'd like me to incorporate?\"\n2.  \"Are there any API rate limits I should be aware of for services used in\n    this workflow that aren't already covered by an existing skill?\"\n3.  \"Are there specific files that provide important scientific context for\n    creating this skill? For example: API documentation, reference papers,\n    example datasets, or domain-specific notes. If so, please share them and I\n    will incorporate their content into the skill's reference materials.\"\n\n### Round 4: Scope and Shape\n\n1.  \"Our workflow covered [X, Y, Z]. Should I distill all of these into the\n    skill, or is there additional functionality that's important to include?\n    Conversely, should any of these be left out?\"\n2.  Determine whether the skill needs any code. If any step involves calling an\n    API, processing data, reading\u002Fwriting files, or computing results, the skill\n    **needs code** and you should default to the CLI pattern. Only use a\n    text-only instruction skill when every step is purely about reasoning,\n    coordinating existing tools, or following a written protocol with no\n    programmatic work at all. Confirm your assessment with the user in plain\n    language:\n    -   If code is needed: \"Some of these steps involve [fetching data from an\n        API \u002F processing files \u002F computing results], so I'll create a helper\n        script that the agent can run for you. The script will have simple\n        commands like `search`, `fetch`, `analyze`, etc. — you won't need to\n        write any code yourself. Does that sound right?\"\n    -   If no code is needed: \"This workflow is entirely about following a set\n        of steps and using existing tools — no new code is needed. I'll write it\n        as a set of clear instructions the agent follows. Does that sound\n        right?\"\n3.  If a helper script will be created: \"I'm thinking the script should have\n    these commands: [proposed commands in plain English, e.g. 'search for\n    proteins', 'fetch results', 'compare sequences']. What would you add or\n    change?\"\n4.  \"What should the skill be called? Proposed name: `[suggestion]`.\"\n\n### Round 5: Testing (Optional)\n\n1.  \"Can you provide a sample query and expected answer that I can use to verify\n    the skill works as intended? For example: 'If I ask [question], the skill\n    should produce [answer].' This is optional but helps me validate the skill\n    during development.\"\n\n### Brainstorming Completion Criteria\n\nYou are ready to move to Phase 2 when you can confidently answer ALL of:\n\n-   [ ] What is the workflow's purpose and scope?\n-   [ ] What are the inputs and outputs?\n-   [ ] Which steps are strict vs. flexible?\n-   [ ] Which existing skills should be referenced?\n-   [ ] What new scripts (if any) are needed?\n-   [ ] What rate limits apply?\n-   [ ] How should errors be handled?\n-   [ ] Does the workflow need any code? (If yes → CLI pattern; if no →\n    instruction-only)\n-   [ ] Where should the skill be installed? (local, global, or custom path)\n-   [ ] Is there a sample query\u002Fanswer for validation?\n\n## Phase 2: Skill Design\n\nProduce a **design document** (as an artifact \u002F implementation plan) and present\nit to the user for approval. The document must include:\n\n1.  **Skill name and description** (following YAML frontmatter rules: name ≤64\n    chars, lowercase + hyphens; description ≤1024 chars).\n2.  **Directory structure** showing all planned files and the **install\n    location** (local, global, or custom — see Rule 7).\n3.  **Existing skills referenced** with rationale for each.\n4.  **New scripts** (if any) with proposed subcommands and arguments.\n5.  **Rate limiting strategy** for any APIs not covered by existing skills.\n6.  **Error handling strategy** per step.\n\n**Wait for explicit user approval before proceeding to Phase 3.**\n\n## Phase 3: Implementation\n\n### Guiding Principles\n\nGeneral guidelines for skill implementation:\n\n-   Use `uv run`, never `python` or `python3`.\n-   Prefer stdlib libraries that come with a default Python 3 installation\n    (`urllib` preferred); Avoid libraries that require extra installation if\n    possible.\n-   Rate limits must be documented and respected in code. Prefer\n    **file-lock–based rate limiting** so that concurrent sub-agents sharing the\n    same machine collectively respect the limit. See other skills in the Science\n    Skills bundle for the canonical cross-process–safe implementation.\n-   Skill output must be \u003C500 lines or redirected to a file. Long output files\n    should be processed programmatically to extract relevant fields.\n-   Hyphens are recommended for the skill name and YAML `name:` field.\n\n### Rule 1: Reuse Existing Skills\n\nWhen the workflow uses functionality covered by an existing installed skill, the\nnew SKILL.md **MUST** reference it by name rather than reimplementing. Include a\n**Dependencies** section in the SKILL.md listing required skills with a brief\nrationale for each.\n\n### Rule 2: Rate Limiting for New APIs\n\nFor any API interaction **not** covered by an existing skill, the generated CLI\nscript **MUST** implement rate limiting. Before writing any rate-limiting code,\n**look up the API's official rate-limit guidelines**: check any documentation\nthe user provided during brainstorming, then search the API's public\ndocumentation online. If no documented rate limit can be found, **default to 1\nrequest per second**. The rate limiting pattern is built directly into the CLI\ntemplate at `references\u002Fcli_script_template.py` — see the `RateLimitError` class\nand the `_request` method of the API client.\n\nKey requirements:\n\n-   Use `time.monotonic()` for timing (not `time.time()`).\n-   Calculate delay from documented rate limits.\n-   Implement retry with exponential backoff for transient errors (5xx).\n-   Raise a dedicated `RateLimitError` when HTTP 429 is received.\n-   Log retry attempts to stderr so the agent can observe progress.\n-   Include the URL and rate-limit value in error messages.\n-   On non-retriable HTTP errors (e.g. 400, 403, 404), read and include the\n    response body in the error message — not just the status code. API response\n    bodies contain actionable details (e.g., \"Invalid parameter\") that enable\n    the agent to self-correct.\n\n### Rule 3: CLI Script Pattern (Default When Code Is Needed)\n\n**This is the default choice.** If **any** step in the workflow involves API\ncalls, data processing, file I\u002FO, computation, or any other programmatic work,\nproduce a multi-command CLI script using `argparse` with subcommands. Follow the\ntemplate in `references\u002Fcli_script_template.py`.\n\nKey requirements:\n\n-   Each major workflow step becomes a subcommand.\n-   All subcommands accept `--output` for writing results to a file.\n-   Use `json.dump` with `indent=2` for JSON output.\n-   Print a success message with the output file path.\n-   Exit with code 1 on errors.\n-   Make arguments like `--limit` **required** (no silent defaults). This forces\n    the agent to specify the value explicitly, preventing it from assuming it\n    retrieved \"all\" results when it was silently capped.\n\n### Rule 4: Default to File Output\n\nAll scripts and workflows **MUST** write output to files, not stdout. Stdout\nshould only contain short status messages (e.g., \"Success! Data written to:\nresults.json\"). This is critical because:\n\n-   API responses can be very large and will truncate in terminal output.\n-   File output is token-efficient — the agent reads only the fields it needs\n    using `jp` or Python one-liners.\n-   Large stdout output wastes context window space.\n\n### Rule 5: Instruction-Only Pattern (Only When No Code Is Needed)\n\nUse this pattern **only** when the workflow requires **zero** programmatic work\n— i.e., every step is purely about orchestration, reasoning, multi-skill\ncoordination, or following a written protocol. If any step needs code (API\ncalls, data processing, file I\u002FO, etc.), use the CLI pattern from Rule 3\ninstead. Produce a SKILL.md with a structured workflow section:\n\n```markdown\n## Workflow\n\n### 1. Step Name\n- Description of what to do\n- Which skill to use and how\n\n### 2. Next Step\n...\n```\n\n### Rule 6: SKILL.md Structure\n\nEvery generated SKILL.md must follow this structure:\n\n```markdown\n---\nname: {skill-name}\ndescription: >-\n  {description}\n---\n\n# {Skill Title}\n\n## Overview\n{Brief description of what the skill does.}\n\n## Dependencies\n{List of required skills, if any.}\n\n## Quick Start\n{Minimal example to get started.}\n\n## Utility Scripts (if CLI-based)\n{Document each subcommand with examples.}\n\n## Workflow (if instruction-only)\n{Numbered steps with clear instructions.}\n\n## Rate Limiting (if applicable)\n{Document rate limits and how they are enforced.}\n\n## Common Mistakes\n{List 2-3 common pitfalls.}\n```\n\n### Rule 7: Skill Placement\n\nSkills can be installed **locally** (project-specific, per-workspace) or\n**globally** (available across all projects). Ask the user if they want to\ninstall locally, globally, or use a custom install path.\n\nSkill paths by agent CLI (local paths relative to project root):\n\n-   **Claude Code**: local `.claude\u002Fskills\u002F`, global `~\u002F.claude\u002Fskills\u002F`\n-   **Codex**: local `.agents\u002Fskills\u002F`, global `~\u002F.agents\u002Fskills\u002F`\n-   **Antigravity 2.0**: local `.agents\u002Fskills\u002F`, global\n    `~\u002F.gemini\u002Fconfig\u002Fskills\u002F`\n-   **Gemini CLI**: local `.gemini\u002Fskills\u002F`, global `~\u002F.gemini\u002Fskills\u002F`\n-   **OpenCode**: local `.opencode\u002Fskills\u002F`, global `~\u002F.config\u002Fopencode\u002Fskills\u002F`\n\nFor other CLIs, find out local and global skill paths by yourself.\n\n## Phase 4: Validation\n\nAfter implementation is complete:\n\n1.  **Test the skill manually** by invoking the agent with a natural-language\n    prompt that should trigger the new skill.\n\n2.  **If a sample query\u002Fanswer was provided** during brainstorming, run it\n    through the skill and verify the output matches expectations.\n","# Workflow-to-Skill Distiller\n\nTurns a completed workflow into a reusable agent skill. Specifically, this skill\nextracts patterns from an interaction or workflow that **already happened** and\npackages them.\n\n> [!CAUTION] **You MUST complete Phase 1 (Brainstorming) before writing any code\n> or SKILL.md content.** Skipping brainstorming produces skills that are either\n> too rigid or too vague. The brainstorming conversation is the most important\n> part of this process.\n\n## Phase 1: Brainstorming (MANDATORY)\n\nHave an **iterative back-and-forth conversation** with the user. Do NOT ask all\nquestions at once. Pick 2-3 relevant questions per round from the bank below,\nrefine your understanding, and ask follow-ups.\n\n### Round 1: Understand the Workflow\n\nStart by summarizing what you observed from the workflow, then ask:\n\n1.  \"Here's my understanding of the workflow: [summary]. Is this accurate? What\n    would you change?\"\n2.  \"What are the expected inputs and outputs for this workflow?\"\n3.  \"How often do you expect to run this workflow? Is it recurring or one-off?\"\n\n### Round 2: Flexibility and Error Handling\n\nFor each step identified in the workflow, determine its rigidity:\n\n1.  \"For [step X], if the primary approach fails (e.g., API down, no results),\n    should the agent: (a) ask you for guidance, (b) try alternative approaches\n    automatically, or (c) fail loudly with an error?\"\n2.  \"Are there any steps where the exact method matters (e.g., must use a\n    specific database), vs. steps where any reasonable approach is fine?\"\n3.  \"Should the skill handle edge cases silently or surface them to the user?\"\n\n### Round 3: Dependencies and Resources\n\nBefore asking these questions, check which of your installed skills overlap with\nthe workflow. If an existing skill from the science bundle covers a step, the\nnew skill **MUST** reference it — do not offer a self-contained option.\n\n1.  \"I noticed the workflow uses functionality covered by [existing skill X,\n    skill Y]. The new skill will reference these rather than reimplementing\n    them. Are there any other tools or skills you'd like me to incorporate?\"\n2.  \"Are there any API rate limits I should be aware of for services used in\n    this workflow that aren't already covered by an existing skill?\"\n3.  \"Are there specific files that provide important scientific context for\n    creating this skill? For example: API documentation, reference papers,\n    example datasets, or domain-specific notes. If so, please share them and I\n    will incorporate their content into the skill's reference materials.\"\n\n### Round 4: Scope and Shape\n\n1.  \"Our workflow covered [X, Y, Z]. Should I distill all of these into the\n    skill, or is there additional functionality that's important to include?\n    Conversely, should any of these be left out?\"\n2.  Determine whether the skill needs any code. If any step involves calling an\n    API, processing data, reading\u002Fwriting files, or computing results, the skill\n    **needs code** and you should default to the CLI pattern. Only use a\n    text-only instruction skill when every step is purely about reasoning,\n    coordinating existing tools, or following a written protocol with no\n    programmatic work at all. Confirm your assessment with the user in plain\n    language:\n    -   If code is needed: \"Some of these steps involve [fetching data from an\n        API \u002F processing files \u002F computing results], so I'll create a helper\n        script that the agent can run for you. The script will have simple\n        commands like `search`, `fetch`, `analyze`, etc. — you won't need to\n        write any code yourself. Does that sound right?\"\n    -   If no code is needed: \"This workflow is entirely about following a set\n        of steps and using existing tools — no new code is needed. I'll write it\n        as a set of clear instructions the agent follows. Does that sound\n        right?\"\n3.  If a helper script will be created: \"I'm thinking the script should have\n    these commands: [proposed commands in plain English, e.g. 'search for\n    proteins', 'fetch results', 'compare sequences']. What would you add or\n    change?\"\n4.  \"What should the skill be called? Proposed name: `[suggestion]`.\"\n\n### Round 5: Testing (Optional)\n\n1.  \"Can you provide a sample query and expected answer that I can use to verify\n    the skill works as intended? For example: 'If I ask [question], the skill\n    should produce [answer].' This is optional but helps me validate the skill\n    during development.\"\n\n### Brainstorming Completion Criteria\n\nYou are ready to move to Phase 2 when you can confidently answer ALL of:\n\n-   [ ] What is the workflow's purpose and scope?\n-   [ ] What are the inputs and outputs?\n-   [ ] Which steps are strict vs. flexible?\n-   [ ] Which existing skills should be referenced?\n-   [ ] What new scripts (if any) are needed?\n-   [ ] What rate limits apply?\n-   [ ] How should errors be handled?\n-   [ ] Does the workflow need any code? (If yes → CLI pattern; if no →\n    instruction-only)\n-   [ ] Where should the skill be installed? (local, global, or custom path)\n-   [ ] Is there a sample query\u002Fanswer for validation?\n\n## Phase 2: Skill Design\n\nProduce a **design document** (as an artifact \u002F implementation plan) and present\nit to the user for approval. The document must include:\n\n1.  **Skill name and description** (following YAML frontmatter rules: name ≤64\n    chars, lowercase + hyphens; description ≤1024 chars).\n2.  **Directory structure** showing all planned files and the **install\n    location** (local, global, or custom — see Rule 7).\n3.  **Existing skills referenced** with rationale for each.\n4.  **New scripts** (if any) with proposed subcommands and arguments.\n5.  **Rate limiting strategy** for any APIs not covered by existing skills.\n6.  **Error handling strategy** per step.\n\n**Wait for explicit user approval before proceeding to Phase 3.**\n\n## Phase 3: Implementation\n\n### Guiding Principles\n\nGeneral guidelines for skill implementation:\n\n-   Use `uv run`, never `python` or `python3`.\n-   Prefer stdlib libraries that come with a default Python 3 installation\n    (`urllib` preferred); Avoid libraries that require extra installation if\n    possible.\n-   Rate limits must be documented and respected in code. Prefer\n    **file-lock–based rate limiting** so that concurrent sub-agents sharing the\n    same machine collectively respect the limit. See other skills in the Science\n    Skills bundle for the canonical cross-process–safe implementation.\n-   Skill output must be \u003C500 lines or redirected to a file. Long output files\n    should be processed programmatically to extract relevant fields.\n-   Hyphens are recommended for the skill name and YAML `name:` field.\n\n### Rule 1: Reuse Existing Skills\n\nWhen the workflow uses functionality covered by an existing installed skill, the\nnew SKILL.md **MUST** reference it by name rather than reimplementing. Include a\n**Dependencies** section in the SKILL.md listing required skills with a brief\nrationale for each.\n\n### Rule 2: Rate Limiting for New APIs\n\nFor any API interaction **not** covered by an existing skill, the generated CLI\nscript **MUST** implement rate limiting. Before writing any rate-limiting code,\n**look up the API's official rate-limit guidelines**: check any documentation\nthe user provided during brainstorming, then search the API's public\ndocumentation online. If no documented rate limit can be found, **default to 1\nrequest per second**. The rate limiting pattern is built directly into the CLI\ntemplate at `references\u002Fcli_script_template.py` — see the `RateLimitError` class\nand the `_request` method of the API client.\n\nKey requirements:\n\n-   Use `time.monotonic()` for timing (not `time.time()`).\n-   Calculate delay from documented rate limits.\n-   Implement retry with exponential backoff for transient errors (5xx).\n-   Raise a dedicated `RateLimitError` when HTTP 429 is received.\n-   Log retry attempts to stderr so the agent can observe progress.\n-   Include the URL and rate-limit value in error messages.\n-   On non-retriable HTTP errors (e.g. 400, 403, 404), read and include the\n    response body in the error message — not just the status code. API response\n    bodies contain actionable details (e.g., \"Invalid parameter\") that enable\n    the agent to self-correct.\n\n### Rule 3: CLI Script Pattern (Default When Code Is Needed)\n\n**This is the default choice.** If **any** step in the workflow involves API\ncalls, data processing, file I\u002FO, computation, or any other programmatic work,\nproduce a multi-command CLI script using `argparse` with subcommands. Follow the\ntemplate in `references\u002Fcli_script_template.py`.\n\nKey requirements:\n\n-   Each major workflow step becomes a subcommand.\n-   All subcommands accept `--output` for writing results to a file.\n-   Use `json.dump` with `indent=2` for JSON output.\n-   Print a success message with the output file path.\n-   Exit with code 1 on errors.\n-   Make arguments like `--limit` **required** (no silent defaults). This forces\n    the agent to specify the value explicitly, preventing it from assuming it\n    retrieved \"all\" results when it was silently capped.\n\n### Rule 4: Default to File Output\n\nAll scripts and workflows **MUST** write output to files, not stdout. Stdout\nshould only contain short status messages (e.g., \"Success! Data written to:\nresults.json\"). This is critical because:\n\n-   API responses can be very large and will truncate in terminal output.\n-   File output is token-efficient — the agent reads only the fields it needs\n    using `jp` or Python one-liners.\n-   Large stdout output wastes context window space.\n\n### Rule 5: Instruction-Only Pattern (Only When No Code Is Needed)\n\nUse this pattern **only** when the workflow requires **zero** programmatic work\n— i.e., every step is purely about orchestration, reasoning, multi-skill\ncoordination, or following a written protocol. If any step needs code (API\ncalls, data processing, file I\u002FO, etc.), use the CLI pattern from Rule 3\ninstead. Produce a SKILL.md with a structured workflow section:\n\n```markdown\n## Workflow\n\n### 1. Step Name\n- Description of what to do\n- Which skill to use and how\n\n### 2. Next Step\n...\n```\n\n### Rule 6: SKILL.md Structure\n\nEvery generated SKILL.md must follow this structure:\n\n```markdown\n---\nname: {skill-name}\ndescription: >-\n  {description}\n---\n\n# {Skill Title}\n\n## Overview\n{Brief description of what the skill does.}\n\n## Dependencies\n{List of required skills, if any.}\n\n## Quick Start\n{Minimal example to get started.}\n\n## Utility Scripts (if CLI-based)\n{Document each subcommand with examples.}\n\n## Workflow (if instruction-only)\n{Numbered steps with clear instructions.}\n\n## Rate Limiting (if applicable)\n{Document rate limits and how they are enforced.}\n\n## Common Mistakes\n{List 2-3 common pitfalls.}\n```\n\n### Rule 7: Skill Placement\n\nSkills can be installed **locally** (project-specific, per-workspace) or\n**globally** (available across all projects). Ask the user if they want to\ninstall locally, globally, or use a custom install path.\n\nSkill paths by agent CLI (local paths relative to project root):\n\n-   **Claude Code**: local `.claude\u002Fskills\u002F`, global `~\u002F.claude\u002Fskills\u002F`\n-   **Codex**: local `.agents\u002Fskills\u002F`, global `~\u002F.agents\u002Fskills\u002F`\n-   **Antigravity 2.0**: local `.agents\u002Fskills\u002F`, global\n    `~\u002F.gemini\u002Fconfig\u002Fskills\u002F`\n-   **Gemini CLI**: local `.gemini\u002Fskills\u002F`, global `~\u002F.gemini\u002Fskills\u002F`\n-   **OpenCode**: local `.opencode\u002Fskills\u002F`, global `~\u002F.config\u002Fopencode\u002Fskills\u002F`\n\nFor other CLIs, find out local and global skill paths by yourself.\n\n## Phase 4: Validation\n\nAfter implementation is complete:\n\n1.  **Test the skill manually** by invoking the agent with a natural-language\n    prompt that should trigger the new skill.\n\n2.  **If a sample query\u002Fanswer was provided** during brainstorming, run it\n    through the skill and verify the output matches expectations.\n",{"name":5,"description":7},"main","e71245365df78e06cd166602744bf2491f79910e","research-skills-hub\u002Fscience-skills\u002Fworkflow_skill_creator",0,"https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS","https:\u002F\u002Fpengqianhan.github.io\u002FAI-Human-Research-OS\u002F","https:\u002F\u002Fgithub.com\u002Fpengqianhan","Ph.D. candidate of The University of Auckland, New Zealand","2026-11-14 18:08:16",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add pengqianhan\u002FAI-Human-Research-OS --skill workflow-skill-creator",{"tool":39,"command":40},"bunx","bunx skills add pengqianhan\u002FAI-Human-Research-OS --skill workflow-skill-creator",{"tool":42,"command":43},"pnpm","pnpm dlx skills add pengqianhan\u002FAI-Human-Research-OS --skill workflow-skill-creator","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fscience-skills\u002Fworkflow_skill_creator\u002FSKILL.md",true,[47],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Fcli_script_template.py","SCRIPT","脚本",6940,"https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fscience-skills\u002Fworkflow_skill_creator\u002Freferences\u002Fcli_script_template.py",[54,61,68,75,82,89],{"skillNo":55,"name":56,"nameZh":6,"description":57,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":60,"publishedAt":19},"01M74MWWVCFZ8HQS8840F1V1FT","uv","Checks whether the uv Python package manager is installed and installs it if missing. Ensures uv is on PATH. Use when another skill requires uv as a prerequisite.","https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fscience-skills\u002Fuv\u002FSKILL.md",[],[],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":19},"01M74MWRH5BTSWZ2C5PCTD5FV9","uniprot-database","Access protein metadata, function, taxonomy, and sequences across UniProtKB, UniParc, and UniRef. Use when searching for proteins, mapping identifiers, or retrieving functional annotations and publications. Don't use for sequence alignment, protein folding, or sequence similarity search (use specialized skills for those tasks).","https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fscience-skills\u002Funiprot_database\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":19},"01M74MWN1D3ER9F1FZPWH8HQC1","unibind-database","Queries the UniBind database for experimentally validated transcription factor (TF) binding sites. Use when retrieving direct TF-DNA interaction datasets, downloading binding site coordinates (BED\u002FFASTA) for local analysis, or listing available datasets by species, cell line, or TF name. Don't use to query specific intervals, locations, genes, motif models or expression data.","https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fscience-skills\u002Funibind_database\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M74MWHHB5QZY3NR2FYA1EK4N","ucsc-conservation-and-tfbs","Fetch Evolutionary Conservation scores (phyloP, phastCons) and Transcription Factor Binding Sites (TFBS) from the UCSC Genome Browser. Use when analyzing whether genomic variants or regions are evolutionarily conserved, functionally important, or bounded by TF regulators across major projects (ENCODE, JASPAR, ReMap).","https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fscience-skills\u002Fucsc_conservation_and_tfbs\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M74MWERYHSY87P21FDMVE6BA","string-database","Query the STRING database for protein-protein interactions (PPIs), functional enrichment, and homology. Use when the user asks about interactions between specific proteins, interaction evidence, confidence scores, protein interaction partners, or pathway enrichments.","https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fscience-skills\u002Fstring_database\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":19},"01M74MWCCEQ7BV6AT5QCHPNHF8","scienceskillscommon","Shared Python package for Science Skills, currently containing http_client -- a unified HTTP client with rate limiting, retries, and exponential backoff. Not a standalone agent skill. Do not invoke directly.","https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fscience-skills\u002Fscienceskillscommon\u002FSKILL.md",[],[]]