[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M74MPT1A9G3P964H3H77NDG2":3,"skill-same:01M74MPT1A9G3P964H3H77NDG2":69},{"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},"01M74MPT1A9G3P964H3H77NDG2","human-cognition-cache",null,"Scaffold human cognition with a project-local, git-trackable four-quadrant cache. Use when (1) initializing, reading, updating, or migrating the cache; (2) completing a task beyond the human's confirmed understanding, where cognition should guide agent autonomy and decision escalation; (3) explaining a verified artifact at the human's level and recording evidence-backed cognitive growth; or (4) finishing any substantive task, where the conversation revealed durable cognition that should be captured without the human having to ask.","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\u002Fopen-paper-skills\u002Fhuman-cognition-cache\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: human-cognition-cache\ndescription: Scaffold human cognition with a project-local, git-trackable four-quadrant cache. Use when (1) initializing, reading, updating, or migrating the cache; (2) completing a task beyond the human's confirmed understanding, where cognition should guide agent autonomy and decision escalation; (3) explaining a verified artifact at the human's level and recording evidence-backed cognitive growth; or (4) finishing any substantive task, where the conversation revealed durable cognition that should be captured without the human having to ask.\n---\n\n# Human Cognition Cache\n\nUse the human's current cognition as a **scaffold**, not a ceiling. Help the\nhuman complete tasks beyond their present knowledge, preserve human authority\nover goals and consequential choices, and turn verified delivery into durable\nunderstanding when the task calls for it.\n\n## Contract\n\nScope:\n- Model the human's cognition in this project: what they can state, know they\n  do not know, reveal through judgment, or may not yet realize.\n- Use that model to choose when to act, ask, explain, or teach.\n- Keep project facts and agent knowledge in their own project artifacts.\n\nInputs:\n- The current request, conversation, and verified task artifacts.\n- The relevant entries in the project-local cognition cache.\n- Human corrections to entries the agent recorded on its own.\n\nOutputs:\n- A maintained Markdown cognition cache with indexed, evidence-labelled entries.\n- For cognition-aware tasks, an explicit allocation of agent-owned work and\n  human-owned decisions.\n- When understanding matters, an artifact-grounded cognitive handoff and any\n  evidence-backed cognition transitions.\n\nLimitations:\n- The cache is an open-world, fallible model. Missing entries mean\n  `unobserved`, not ignorance.\n- Agent expertise is also fallible; verify technical claims through source,\n  documentation, tests, or experiments.\n- Cognitive growth requires evidence within a stated scope. One successful\n  action does not establish broad mastery.\n\n## Core Model\n\n- `known_knowns.md`: cognition the human can state or demonstrate clearly.\n- `known_unknowns.md`: questions or gaps the human recognizes.\n- `unknown_knowns.md`: implicit criteria or knowledge revealed through\n  reactions, corrections, examples, or repeated choices.\n- `unknown_unknowns.md`: candidate blind spots, stored as questions or\n  hypotheses rather than facts.\n- `unobserved`: a runtime classification for relevant concepts with no adequate\n  evidence. Do not persist it merely because the cache is silent.\n\nKeep the same stable cognition ID when an entry moves between quadrants.\n\n## Autonomous Capture\n\nKeeping the cache current is the agent's job, not the human's. Waiting to be\ninvoked costs continuity and starves the two quadrants the human cannot\nself-report: nobody can state their own implicit criteria or blind spots.\n\nCapture at task boundaries, not every turn. When a substantive task finishes,\nreview the whole conversation for what it revealed about the human's cognition\nand record it in one pass. A single turn rarely carries enough evidence, and\nper-turn capture manufactures low-confidence noise.\n\nWrite without asking, then say so. Record the entry, and close the reply with\none line naming the ID, such as `Recorded cognition cog-20260724-001`. Do not\ninterrupt the task to request permission, do not open a pending queue, and do\nnot stage a review ritual. The human's word alone is enough to correct or delete\nany entry; version control carries the rest of the audit trail.\n\nPersist any non-private cognition. There is no topic filter: engineering,\ntooling, and workflow cognition shape agent routing at least as much as subject\nknowledge does. The only gates are the privacy guardrails below and the\nrequirement that the observation be durable and human-scoped.\n\n**Merge before appending.** Scan the target quadrant's `Active Index` first.\nWhen an existing entry already covers the topic, update it — append evidence,\nadjust confidence, widen or narrow scope, refresh `last_verified` — instead of\ncreating a new ID. This is a hard prerequisite, not a preference: silent capture\nhas no size cap, so unchecked duplicates are its main failure mode and the human\ncannot see them accumulating.\n\nLet observed evidence set confidence. Mark autonomous entries `source:\ninferred`, but take `evidence_type` and `confidence` from what actually\nhappened: a demonstration witnessed in conversation is real evidence and should\nnot be discounted for lacking confirmation. Overstating cognition is the more\nharmful error, because it makes the agent explain too little and leaves the\nhuman silently behind, so `known_knowns` requires `explanation`-level evidence\nor stronger. Inference alone never promotes an entry there.\n\nRecord blind spots, then raise them. Write `unknown_unknowns` entries as\nquestions or hypotheses, never diagnoses, and surface the relevant one when a\nlater task touches it. A blind spot the human is never told about has no value\nto them.\n\nSee [references\u002Fcache-schema.md](references\u002Fcache-schema.md) for decay,\narchiving, and entry fields.\n\n## Branches and Context Pointers\n\n- **Cache maintenance**: when initializing, migrating, writing, or moving\n  entries, read [references\u002Fcache-schema.md](references\u002Fcache-schema.md) before\n  editing. It is the single source of truth for layout, fields, indexes, and\n  transition stubs.\n- **Beyond-cognition execution**: when the task may exceed the human's confirmed\n  understanding or creates consequential cognitive debt, read\n  [references\u002Fcognition-aware-execution.md](references\u002Fcognition-aware-execution.md)\n  before deciding what to ask or execute.\n- **Cognitive handoff**: when the human needs to understand, operate, review,\n  maintain, or extend a delivered artifact, read\n  [references\u002Fartifact-grounded-teaching.md](references\u002Fartifact-grounded-teaching.md)\n  before explaining or updating cognition levels.\n\nLoad only the references reached by the active branch.\n\n## Steps\n\n### 1. Bound the relevant cognition\n\nLocate the cache, skim `index.md`, then skim the frontmatter and `Active Index`\nof all four quadrant files. Build a small relevance set from concepts and\ndecisions that materially affect the current task. Read full entries only for\nthat set. When initialization is the active branch and no cache exists, record\nthe absence and continue to the schema reference.\n\nClassify each relevant concept as confirmed known, declared unknown, inferred\nimplicit, suspected blind spot, or unobserved. Preserve the entry's evidence,\nscope, confidence, and freshness.\n\nComplete when every material concept has a supported classification or is\nexplicitly marked unobserved.\n\n### 2. Select and run the branch\n\nFor cache maintenance, apply the schema reference and preserve stable IDs.\n\nFor beyond-cognition execution, treat a gap as a reason for greater agent\ninitiative, not a reason to shrink the goal. Assign technical investigation and\nreversible implementation choices to the agent. Escalate goals, values,\npermissions, budgets, irreversible actions, and material risk choices to the\nhuman, supplying the minimum background needed to decide.\n\nFor a cognitive handoff, choose the target capability and explanation depth,\nthen teach from the verified artifact, its evidence, and the human's existing\nknown knowns.\n\nComplete when each material unknown has an owner, each required human decision\nis informed, and the active branch's completion criterion is met.\n\n### 3. Deliver both required outcomes\n\nVerify the task artifact independently of the cognition cache. When human\nunderstanding is part of the task, also deliver the cognitive handoff and\ncollect proportionate evidence of understanding through explanation,\noperation, diagnosis, modification, or transfer.\n\nReport these fields separately:\n\n- `artifact_status`: verified, partially verified, or unverified;\n- `target_capability`: awareness, operator, interpreter, modifier, builder, or\n  not required;\n- `cognition_evidence`: observed evidence or `not assessed`;\n- `cognitive_status`: complete, partial, or not assessed;\n- `remaining_debt`: unresolved assumptions and accepted cognitive debt.\n\nComplete when every field is reported and cognitive completion is claimed only\nat the scope supported by observed evidence.\n\n### 4. Persist durable cognition\n\nRun this step at the end of every substantive task, whether or not the human\nasked for it, following the Autonomous Capture rules above. Persist only\nhuman-scoped, durable cognition that will improve future collaboration, and\npreserve the narrowest scope supported by the evidence. A read-only or\ninconclusive task may correctly produce no cache edit.\n\nAlso apply the decay rules while you are here: refresh `last_verified` on any\nentry this task re-confirmed, and mark or archive entries the schema has aged\nout.\n\nComplete when every persisted entry has a stable ID, source, confidence,\nevidence, dates, status, and current index entry; every move has a transition\nstub; no new entry duplicates an existing one; every write is named in the\nreply's closing line; and any high-sensitivity claim was withheld rather than\nrecorded.\n\n## Boundaries\n\n- Store source facts, API constraints, implementation details, and agent\n  uncertainty in project documentation, plans, ADRs, or task notes.\n- Store stable personal context and collaboration preferences in the project's\n  human profile when one exists; use this cache for cognition state.\n- Treat behavioral signals such as pauses, cursor movement, or repeated edits as\n  low-confidence candidate evidence until corroborated.\n- Preserve human choice over whether a low-risk delivery also needs teaching.\n- Before publishing or pushing a tracked cache, warn that it may expose the\n  human's cognition state.\n\n## Privacy Guardrails\n\n- Never record credentials, recovery material, government IDs, bank details,\n  medical records, or other high-sensitivity personal data.\n- Record cognition claims — including capability and long-term ones — without\n  asking, and disclose each write in the reply's closing line. Continuity of the\n  conversation outweighs a confirmation ritual for cognition the human can\n  revoke with one sentence. This exemption covers cognition only; it never\n  licenses recording the high-sensitivity data listed above, which is withheld\n  rather than confirmed.\n- Ask before recording identifying personal claims that are not cognition, such\n  as location, employer, or affiliation.\n- Mark agent-derived observations as `source: inferred` and keep confidence\n  conservative.\n- Record candidate blind spots as questions or hypotheses, never diagnoses.\n","# Human Cognition Cache\n\nUse the human's current cognition as a **scaffold**, not a ceiling. Help the\nhuman complete tasks beyond their present knowledge, preserve human authority\nover goals and consequential choices, and turn verified delivery into durable\nunderstanding when the task calls for it.\n\n## Contract\n\nScope:\n- Model the human's cognition in this project: what they can state, know they\n  do not know, reveal through judgment, or may not yet realize.\n- Use that model to choose when to act, ask, explain, or teach.\n- Keep project facts and agent knowledge in their own project artifacts.\n\nInputs:\n- The current request, conversation, and verified task artifacts.\n- The relevant entries in the project-local cognition cache.\n- Human corrections to entries the agent recorded on its own.\n\nOutputs:\n- A maintained Markdown cognition cache with indexed, evidence-labelled entries.\n- For cognition-aware tasks, an explicit allocation of agent-owned work and\n  human-owned decisions.\n- When understanding matters, an artifact-grounded cognitive handoff and any\n  evidence-backed cognition transitions.\n\nLimitations:\n- The cache is an open-world, fallible model. Missing entries mean\n  `unobserved`, not ignorance.\n- Agent expertise is also fallible; verify technical claims through source,\n  documentation, tests, or experiments.\n- Cognitive growth requires evidence within a stated scope. One successful\n  action does not establish broad mastery.\n\n## Core Model\n\n- `known_knowns.md`: cognition the human can state or demonstrate clearly.\n- `known_unknowns.md`: questions or gaps the human recognizes.\n- `unknown_knowns.md`: implicit criteria or knowledge revealed through\n  reactions, corrections, examples, or repeated choices.\n- `unknown_unknowns.md`: candidate blind spots, stored as questions or\n  hypotheses rather than facts.\n- `unobserved`: a runtime classification for relevant concepts with no adequate\n  evidence. Do not persist it merely because the cache is silent.\n\nKeep the same stable cognition ID when an entry moves between quadrants.\n\n## Autonomous Capture\n\nKeeping the cache current is the agent's job, not the human's. Waiting to be\ninvoked costs continuity and starves the two quadrants the human cannot\nself-report: nobody can state their own implicit criteria or blind spots.\n\nCapture at task boundaries, not every turn. When a substantive task finishes,\nreview the whole conversation for what it revealed about the human's cognition\nand record it in one pass. A single turn rarely carries enough evidence, and\nper-turn capture manufactures low-confidence noise.\n\nWrite without asking, then say so. Record the entry, and close the reply with\none line naming the ID, such as `Recorded cognition cog-20260724-001`. Do not\ninterrupt the task to request permission, do not open a pending queue, and do\nnot stage a review ritual. The human's word alone is enough to correct or delete\nany entry; version control carries the rest of the audit trail.\n\nPersist any non-private cognition. There is no topic filter: engineering,\ntooling, and workflow cognition shape agent routing at least as much as subject\nknowledge does. The only gates are the privacy guardrails below and the\nrequirement that the observation be durable and human-scoped.\n\n**Merge before appending.** Scan the target quadrant's `Active Index` first.\nWhen an existing entry already covers the topic, update it — append evidence,\nadjust confidence, widen or narrow scope, refresh `last_verified` — instead of\ncreating a new ID. This is a hard prerequisite, not a preference: silent capture\nhas no size cap, so unchecked duplicates are its main failure mode and the human\ncannot see them accumulating.\n\nLet observed evidence set confidence. Mark autonomous entries `source:\ninferred`, but take `evidence_type` and `confidence` from what actually\nhappened: a demonstration witnessed in conversation is real evidence and should\nnot be discounted for lacking confirmation. Overstating cognition is the more\nharmful error, because it makes the agent explain too little and leaves the\nhuman silently behind, so `known_knowns` requires `explanation`-level evidence\nor stronger. Inference alone never promotes an entry there.\n\nRecord blind spots, then raise them. Write `unknown_unknowns` entries as\nquestions or hypotheses, never diagnoses, and surface the relevant one when a\nlater task touches it. A blind spot the human is never told about has no value\nto them.\n\nSee [references\u002Fcache-schema.md](references\u002Fcache-schema.md) for decay,\narchiving, and entry fields.\n\n## Branches and Context Pointers\n\n- **Cache maintenance**: when initializing, migrating, writing, or moving\n  entries, read [references\u002Fcache-schema.md](references\u002Fcache-schema.md) before\n  editing. It is the single source of truth for layout, fields, indexes, and\n  transition stubs.\n- **Beyond-cognition execution**: when the task may exceed the human's confirmed\n  understanding or creates consequential cognitive debt, read\n  [references\u002Fcognition-aware-execution.md](references\u002Fcognition-aware-execution.md)\n  before deciding what to ask or execute.\n- **Cognitive handoff**: when the human needs to understand, operate, review,\n  maintain, or extend a delivered artifact, read\n  [references\u002Fartifact-grounded-teaching.md](references\u002Fartifact-grounded-teaching.md)\n  before explaining or updating cognition levels.\n\nLoad only the references reached by the active branch.\n\n## Steps\n\n### 1. Bound the relevant cognition\n\nLocate the cache, skim `index.md`, then skim the frontmatter and `Active Index`\nof all four quadrant files. Build a small relevance set from concepts and\ndecisions that materially affect the current task. Read full entries only for\nthat set. When initialization is the active branch and no cache exists, record\nthe absence and continue to the schema reference.\n\nClassify each relevant concept as confirmed known, declared unknown, inferred\nimplicit, suspected blind spot, or unobserved. Preserve the entry's evidence,\nscope, confidence, and freshness.\n\nComplete when every material concept has a supported classification or is\nexplicitly marked unobserved.\n\n### 2. Select and run the branch\n\nFor cache maintenance, apply the schema reference and preserve stable IDs.\n\nFor beyond-cognition execution, treat a gap as a reason for greater agent\ninitiative, not a reason to shrink the goal. Assign technical investigation and\nreversible implementation choices to the agent. Escalate goals, values,\npermissions, budgets, irreversible actions, and material risk choices to the\nhuman, supplying the minimum background needed to decide.\n\nFor a cognitive handoff, choose the target capability and explanation depth,\nthen teach from the verified artifact, its evidence, and the human's existing\nknown knowns.\n\nComplete when each material unknown has an owner, each required human decision\nis informed, and the active branch's completion criterion is met.\n\n### 3. Deliver both required outcomes\n\nVerify the task artifact independently of the cognition cache. When human\nunderstanding is part of the task, also deliver the cognitive handoff and\ncollect proportionate evidence of understanding through explanation,\noperation, diagnosis, modification, or transfer.\n\nReport these fields separately:\n\n- `artifact_status`: verified, partially verified, or unverified;\n- `target_capability`: awareness, operator, interpreter, modifier, builder, or\n  not required;\n- `cognition_evidence`: observed evidence or `not assessed`;\n- `cognitive_status`: complete, partial, or not assessed;\n- `remaining_debt`: unresolved assumptions and accepted cognitive debt.\n\nComplete when every field is reported and cognitive completion is claimed only\nat the scope supported by observed evidence.\n\n### 4. Persist durable cognition\n\nRun this step at the end of every substantive task, whether or not the human\nasked for it, following the Autonomous Capture rules above. Persist only\nhuman-scoped, durable cognition that will improve future collaboration, and\npreserve the narrowest scope supported by the evidence. A read-only or\ninconclusive task may correctly produce no cache edit.\n\nAlso apply the decay rules while you are here: refresh `last_verified` on any\nentry this task re-confirmed, and mark or archive entries the schema has aged\nout.\n\nComplete when every persisted entry has a stable ID, source, confidence,\nevidence, dates, status, and current index entry; every move has a transition\nstub; no new entry duplicates an existing one; every write is named in the\nreply's closing line; and any high-sensitivity claim was withheld rather than\nrecorded.\n\n## Boundaries\n\n- Store source facts, API constraints, implementation details, and agent\n  uncertainty in project documentation, plans, ADRs, or task notes.\n- Store stable personal context and collaboration preferences in the project's\n  human profile when one exists; use this cache for cognition state.\n- Treat behavioral signals such as pauses, cursor movement, or repeated edits as\n  low-confidence candidate evidence until corroborated.\n- Preserve human choice over whether a low-risk delivery also needs teaching.\n- Before publishing or pushing a tracked cache, warn that it may expose the\n  human's cognition state.\n\n## Privacy Guardrails\n\n- Never record credentials, recovery material, government IDs, bank details,\n  medical records, or other high-sensitivity personal data.\n- Record cognition claims — including capability and long-term ones — without\n  asking, and disclose each write in the reply's closing line. Continuity of the\n  conversation outweighs a confirmation ritual for cognition the human can\n  revoke with one sentence. This exemption covers cognition only; it never\n  licenses recording the high-sensitivity data listed above, which is withheld\n  rather than confirmed.\n- Ask before recording identifying personal claims that are not cognition, such\n  as location, employer, or affiliation.\n- Mark agent-derived observations as `source: inferred` and keep confidence\n  conservative.\n- Record candidate blind spots as questions or hypotheses, never diagnoses.\n",{"name":5,"description":7},"main","e71245365df78e06cd166602744bf2491f79910e","research-skills-hub\u002Fopen-paper-skills\u002Fhuman-cognition-cache",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 human-cognition-cache",{"tool":39,"command":40},"bunx","bunx skills add pengqianhan\u002FAI-Human-Research-OS --skill human-cognition-cache",{"tool":42,"command":43},"pnpm","pnpm dlx skills add pengqianhan\u002FAI-Human-Research-OS --skill human-cognition-cache","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fopen-paper-skills\u002Fhuman-cognition-cache\u002FSKILL.md",true,[47,53,57,61,65],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"agents\u002Fopenai.yaml","REFERENCE","参考文档",273,"https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fopen-paper-skills\u002Fhuman-cognition-cache\u002Fagents\u002Fopenai.yaml",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"README.md",6093,"https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fopen-paper-skills\u002Fhuman-cognition-cache\u002FREADME.md",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"references\u002Fartifact-grounded-teaching.md",3659,"https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fopen-paper-skills\u002Fhuman-cognition-cache\u002Freferences\u002Fartifact-grounded-teaching.md",{"path":62,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":63,"sourceHtmlUrl":64},"references\u002Fcache-schema.md",6985,"https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fopen-paper-skills\u002Fhuman-cognition-cache\u002Freferences\u002Fcache-schema.md",{"path":66,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":67,"sourceHtmlUrl":68},"references\u002Fcognition-aware-execution.md",3902,"https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fopen-paper-skills\u002Fhuman-cognition-cache\u002Freferences\u002Fcognition-aware-execution.md",[70,77,84,91,98,105],{"skillNo":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M74MWYTHT498SVMN823896T7","workflow-skill-creator","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).","https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fscience-skills\u002Fworkflow_skill_creator\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"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":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"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":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"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":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"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":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":109,"categories":110,"keywords":111,"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",[],[]]