[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M74MQBJ4D46KABZ5ZW2CVZBK":3,"skill-same:01M74MQBJ4D46KABZ5ZW2CVZBK":57},{"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},"01M74MQBJ4D46KABZ5ZW2CVZBK","paper-search",null,"Search and read research papers across alphaXiv, OpenAlex, bioRxiv, PubMed, Hugging Face Papers, Papers with Code, and arXiv, honouring the human's per-source on\u002Foff switches and full-text read order. Use when finding papers, related work, prior art, or code for a method; when reading a paper's full text or record from its id; or when the human asks to switch a source on or off or reorder where papers are read from.","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\u002Fpaper-search\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: paper-search\ndescription: Search and read research papers across alphaXiv, OpenAlex, bioRxiv, PubMed, Hugging Face Papers, Papers with Code, and arXiv, honouring the human's per-source on\u002Foff switches and full-text read order. Use when finding papers, related work, prior art, or code for a method; when reading a paper's full text or record from its id; or when the human asks to switch a source on or off or reorder where papers are read from.\n---\n\n# Paper search\n\n`scripts\u002Fpaper_search.py` in this skill's directory (`$PS` below) searches every\nliterature source the human has switched on, in parallel, and returns one JSON\nshape; `fetch` reads one paper through the human's read order. The switches and\nthe read order live in the repository's\n[`memory\u002Fpaper-sources.json`](..\u002F..\u002F..\u002Fmemory\u002Fpaper-sources.json); the human\nflips the switches from the **Papers** button in os-ui's agent windows.\n\nA source that is off is off on every route. Reach literature through `$PS`, and\nleave a switched-off source alone through other routes as well: MCP connectors,\nother skills, and web fetches. When the evidence needs a source that is off,\nsay so and let the human decide.\n\n## Search\n\nRun from anywhere inside the repository; the script finds the switch file by\nwalking up from the working directory.\n\n```bash\nuv run $PS sources                                   # which sources are on\nuv run $PS search \"sparse autoencoder feature splitting\"\nuv run $PS search \"KV cache compression\" --source alphaxiv --source huggingface --published-after 2025-01-01\nuv run $PS search \"ImageNet top-1\" --mode keyword --limit 5\n```\n\nWithout `--source`, every source that is on is searched. Pick the sources that\nfit the field: machine learning reads best through `alphaxiv`, `huggingface`,\nand `pwc`; journal and cross-disciplinary work through `openalex`; biology\nthrough `biorxiv` and `pubmed`; an exact title, author, or arXiv category\nthrough `arxiv` with a fielded query such as `ti:\"context language models\"`.\n\n| Source | Covers | Extra fields |\n|---|---|---|\n| `alphaxiv` | arXiv full text: CS, math, physics, stats | `votes`, full-text `snippets` |\n| `openalex` | all disciplines, journals and conferences | `citations` |\n| `biorxiv` | biology preprints, via OpenAlex's bioRxiv index | `citations` |\n| `pubmed` | biomedical and life-science journals | `doi`, `venue` |\n| `huggingface` | arXiv ML papers on Hugging Face Papers | `votes` (upvotes), `code` (GitHub) |\n| `pwc` | Papers with Code catalog, through its MCP server | `citations`, `code_repos`, `official_code`; no abstract |\n| `arxiv` | all of arXiv, through its own API; `ti:` `au:` `abs:` `cat:` queries pass through | `doi`, `venue` (journal reference) |\n\nControls, applied to every source that supports them:\n\n- `--mode semantic` (default) or `keyword`: alphaXiv embedding vs full-text\n  keyword search, PwC hybrid vs keyword, arXiv all-words vs exact phrase. Use\n  `keyword` for exact method names, acronyms, benchmarks, and title phrases.\n  Other sources ignore it.\n- `--published-after` \u002F `--published-before YYYY-MM-DD`: inclusive window.\n  Supply one only when the question asks for a period.\n- `--limit N`: results per source (default 10; PwC caps at 25).\n\n## Read the result\n\nThe script prints `{query, searched, disabled, errors, results}`. Each result\nhas `source`, `id`, `title`, `url`, and, when its source provides them,\n`abstract`, `publication_date`, `authors`, and the extra fields above. `id` is\nan arXiv id, a DOI, an OpenAlex `W…` id, a PMID, or a PwC catalog id.\n\n- Results interleave the sources rank by rank. A paper several sources found\n  appears once, under the first source, with the others in `also_in`; it keeps\n  that source's `votes` and takes missing fields such as `code` from the rest.\n- `errors` names each source that failed. Report a failed source as unsearched\n  rather than as evidence that no such literature exists; the same holds for\n  an empty result. Exit 0 means at least one source answered, 1 means every\n  searched source failed, 2 means invalid use, including asking for a source\n  that is off.\n- Cite only papers the results or a follow-up read actually returned.\n\nKnown source behaviour:\n\n- PwC allows 10 hybrid searches per minute per IP; bursts need `--mode keyword`.\n  `PWC_MCP_URL` points at another server, such as a local `pwc-mcp`.\n- Hugging Face has no date filter, so a window is applied to a wider pool\n  afterwards and can return fewer results than `--limit`.\n- A PubMed window matches print or electronic dates, while `publication_date`\n  shows the electronic one, so a hit can predate the window.\n- arXiv's terms allow one request every three seconds; the script waits out\n  the gap, so back-to-back arXiv calls are slow rather than refused.\n\n## Keys\n\nEvery user brings their own optional keys: `OPENALEX_API_KEY` (OpenAlex and\nbioRxiv), `NCBI_API_KEY` (PubMed), and `HF_TOKEN` (Hugging Face). The script\nreads them from the environment, else from the repository's gitignored `.env`\n(template: [`.env.example`](..\u002F..\u002F..\u002F.env.example)); `uv run $PS keys` shows\nwhich are set. alphaXiv and PwC need none.\n\nA rate-limited source (`HTTP 429`) comes back with the key that lifts the\nlimit. Relay that next step to the human: they paste the key under **Keys** in\nos-ui's Papers panel, which runs `keys --set NAME` with the value on stdin, or\nedit `.env` themselves. Keep key values out of commits, prompts, and logs.\n\n## Read a paper\n\n`fetch` takes an arXiv id or URL, a DOI, an OpenAlex `W…` id, or a PubMed id\nand routes it by shape:\n\n```bash\nuv run $PS fetch 2609.37725                    # full text through the read order\nuv run $PS fetch 2609.37725 --via arxiv        # one source only\nuv run $PS fetch 2609.37725 --report           # alphaXiv's ~10 KB overview, for skimming\nuv run $PS fetch 10.1038\u002Fnature14539           # a DOI or PMID: its record, no full text\n```\n\nAn arXiv paper is read from each source in the read order (default\n`huggingface`, `alphaxiv`, `arxiv`), skipping switched-off sources, until one\nreturns a complete text: at least 5,000 characters with a heading such as\nIntroduction or Method. A cut-short text, such as the lone figure caption\nHugging Face serves for some fresh submissions, falls through to the next\nsource. The text goes to a file (`--out`, default under the system temp\ndirectory); stdout is a JSON report with `tried` (each source's outcome),\n`source`, `url`, `file`, `chars`, and `metadata` from arXiv, else Hugging Face.\nRead the file in parts and cite the `source` and `url` the report names.\n\nExit 1 means no source returned a complete text: report the `tried` outcomes\nand the `pdf` link, and let the human decide; `--via` or `--report` through a\nswitched-off source exits 2. A DOI, OpenAlex id, or PMID returns its record\nfrom OpenAlex, bioRxiv, or PubMed under that source's switch; follow its\n`url` or `doi` for the text.\n\nWhen the harness has a logged-in alphaXiv connector (Claude Code's, for\ninstance), use it to put questions to a paper's PDF. It belongs to the\n`alphaxiv` source, so it follows that switch.\n\n## Switch a source\n\nbioRxiv and PubMed start switched off and every other source starts on; a\nsource the switch file leaves out keeps that default. Change a switch only on\nthe human's word, since the switches are the human's choice:\n\n```bash\nuv run $PS sources --disable pubmed\nuv run $PS sources --enable pubmed --json\n```\n\nThe read order is the human's choice too: `sources --read-order\nalphaxiv,huggingface,arxiv` sets it, and leaving a source out of it keeps\n`fetch` from reading through it while search still uses it.\n\nBoth rewrite `memory\u002Fpaper-sources.json`, a tracked file; commit it with the\nchange that motivated it.\n\n_Adapted from [alphaXiv OpenResearch](https:\u002F\u002Fgithub.com\u002FalphaXiv\u002FOpenResearch)\n(MIT): its `orx discover` source switches and shared result shape, and its\n`orx paper` id routing and alphaXiv reads, are reimplemented here, and Hugging\nFace Papers and the\n[Papers with Code MCP server](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fpwc-cli#mcp-server)\nare added as sources. The arXiv source reimplements the query syntax and\nrate limit of\n[science-skills' `literature_search_arxiv`](https:\u002F\u002Fgithub.com\u002Fgoogle-deepmind\u002Fscience-skills)\n(Apache-2.0)._\n","# Paper search\n\n`scripts\u002Fpaper_search.py` in this skill's directory (`$PS` below) searches every\nliterature source the human has switched on, in parallel, and returns one JSON\nshape; `fetch` reads one paper through the human's read order. The switches and\nthe read order live in the repository's\n[`memory\u002Fpaper-sources.json`](..\u002F..\u002F..\u002Fmemory\u002Fpaper-sources.json); the human\nflips the switches from the **Papers** button in os-ui's agent windows.\n\nA source that is off is off on every route. Reach literature through `$PS`, and\nleave a switched-off source alone through other routes as well: MCP connectors,\nother skills, and web fetches. When the evidence needs a source that is off,\nsay so and let the human decide.\n\n## Search\n\nRun from anywhere inside the repository; the script finds the switch file by\nwalking up from the working directory.\n\n```bash\nuv run $PS sources                                   # which sources are on\nuv run $PS search \"sparse autoencoder feature splitting\"\nuv run $PS search \"KV cache compression\" --source alphaxiv --source huggingface --published-after 2025-01-01\nuv run $PS search \"ImageNet top-1\" --mode keyword --limit 5\n```\n\nWithout `--source`, every source that is on is searched. Pick the sources that\nfit the field: machine learning reads best through `alphaxiv`, `huggingface`,\nand `pwc`; journal and cross-disciplinary work through `openalex`; biology\nthrough `biorxiv` and `pubmed`; an exact title, author, or arXiv category\nthrough `arxiv` with a fielded query such as `ti:\"context language models\"`.\n\n| Source | Covers | Extra fields |\n|---|---|---|\n| `alphaxiv` | arXiv full text: CS, math, physics, stats | `votes`, full-text `snippets` |\n| `openalex` | all disciplines, journals and conferences | `citations` |\n| `biorxiv` | biology preprints, via OpenAlex's bioRxiv index | `citations` |\n| `pubmed` | biomedical and life-science journals | `doi`, `venue` |\n| `huggingface` | arXiv ML papers on Hugging Face Papers | `votes` (upvotes), `code` (GitHub) |\n| `pwc` | Papers with Code catalog, through its MCP server | `citations`, `code_repos`, `official_code`; no abstract |\n| `arxiv` | all of arXiv, through its own API; `ti:` `au:` `abs:` `cat:` queries pass through | `doi`, `venue` (journal reference) |\n\nControls, applied to every source that supports them:\n\n- `--mode semantic` (default) or `keyword`: alphaXiv embedding vs full-text\n  keyword search, PwC hybrid vs keyword, arXiv all-words vs exact phrase. Use\n  `keyword` for exact method names, acronyms, benchmarks, and title phrases.\n  Other sources ignore it.\n- `--published-after` \u002F `--published-before YYYY-MM-DD`: inclusive window.\n  Supply one only when the question asks for a period.\n- `--limit N`: results per source (default 10; PwC caps at 25).\n\n## Read the result\n\nThe script prints `{query, searched, disabled, errors, results}`. Each result\nhas `source`, `id`, `title`, `url`, and, when its source provides them,\n`abstract`, `publication_date`, `authors`, and the extra fields above. `id` is\nan arXiv id, a DOI, an OpenAlex `W…` id, a PMID, or a PwC catalog id.\n\n- Results interleave the sources rank by rank. A paper several sources found\n  appears once, under the first source, with the others in `also_in`; it keeps\n  that source's `votes` and takes missing fields such as `code` from the rest.\n- `errors` names each source that failed. Report a failed source as unsearched\n  rather than as evidence that no such literature exists; the same holds for\n  an empty result. Exit 0 means at least one source answered, 1 means every\n  searched source failed, 2 means invalid use, including asking for a source\n  that is off.\n- Cite only papers the results or a follow-up read actually returned.\n\nKnown source behaviour:\n\n- PwC allows 10 hybrid searches per minute per IP; bursts need `--mode keyword`.\n  `PWC_MCP_URL` points at another server, such as a local `pwc-mcp`.\n- Hugging Face has no date filter, so a window is applied to a wider pool\n  afterwards and can return fewer results than `--limit`.\n- A PubMed window matches print or electronic dates, while `publication_date`\n  shows the electronic one, so a hit can predate the window.\n- arXiv's terms allow one request every three seconds; the script waits out\n  the gap, so back-to-back arXiv calls are slow rather than refused.\n\n## Keys\n\nEvery user brings their own optional keys: `OPENALEX_API_KEY` (OpenAlex and\nbioRxiv), `NCBI_API_KEY` (PubMed), and `HF_TOKEN` (Hugging Face). The script\nreads them from the environment, else from the repository's gitignored `.env`\n(template: [`.env.example`](..\u002F..\u002F..\u002F.env.example)); `uv run $PS keys` shows\nwhich are set. alphaXiv and PwC need none.\n\nA rate-limited source (`HTTP 429`) comes back with the key that lifts the\nlimit. Relay that next step to the human: they paste the key under **Keys** in\nos-ui's Papers panel, which runs `keys --set NAME` with the value on stdin, or\nedit `.env` themselves. Keep key values out of commits, prompts, and logs.\n\n## Read a paper\n\n`fetch` takes an arXiv id or URL, a DOI, an OpenAlex `W…` id, or a PubMed id\nand routes it by shape:\n\n```bash\nuv run $PS fetch 2609.37725                    # full text through the read order\nuv run $PS fetch 2609.37725 --via arxiv        # one source only\nuv run $PS fetch 2609.37725 --report           # alphaXiv's ~10 KB overview, for skimming\nuv run $PS fetch 10.1038\u002Fnature14539           # a DOI or PMID: its record, no full text\n```\n\nAn arXiv paper is read from each source in the read order (default\n`huggingface`, `alphaxiv`, `arxiv`), skipping switched-off sources, until one\nreturns a complete text: at least 5,000 characters with a heading such as\nIntroduction or Method. A cut-short text, such as the lone figure caption\nHugging Face serves for some fresh submissions, falls through to the next\nsource. The text goes to a file (`--out`, default under the system temp\ndirectory); stdout is a JSON report with `tried` (each source's outcome),\n`source`, `url`, `file`, `chars`, and `metadata` from arXiv, else Hugging Face.\nRead the file in parts and cite the `source` and `url` the report names.\n\nExit 1 means no source returned a complete text: report the `tried` outcomes\nand the `pdf` link, and let the human decide; `--via` or `--report` through a\nswitched-off source exits 2. A DOI, OpenAlex id, or PMID returns its record\nfrom OpenAlex, bioRxiv, or PubMed under that source's switch; follow its\n`url` or `doi` for the text.\n\nWhen the harness has a logged-in alphaXiv connector (Claude Code's, for\ninstance), use it to put questions to a paper's PDF. It belongs to the\n`alphaxiv` source, so it follows that switch.\n\n## Switch a source\n\nbioRxiv and PubMed start switched off and every other source starts on; a\nsource the switch file leaves out keeps that default. Change a switch only on\nthe human's word, since the switches are the human's choice:\n\n```bash\nuv run $PS sources --disable pubmed\nuv run $PS sources --enable pubmed --json\n```\n\nThe read order is the human's choice too: `sources --read-order\nalphaxiv,huggingface,arxiv` sets it, and leaving a source out of it keeps\n`fetch` from reading through it while search still uses it.\n\nBoth rewrite `memory\u002Fpaper-sources.json`, a tracked file; commit it with the\nchange that motivated it.\n\n_Adapted from [alphaXiv OpenResearch](https:\u002F\u002Fgithub.com\u002FalphaXiv\u002FOpenResearch)\n(MIT): its `orx discover` source switches and shared result shape, and its\n`orx paper` id routing and alphaXiv reads, are reimplemented here, and Hugging\nFace Papers and the\n[Papers with Code MCP server](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fpwc-cli#mcp-server)\nare added as sources. The arXiv source reimplements the query syntax and\nrate limit of\n[science-skills' `literature_search_arxiv`](https:\u002F\u002Fgithub.com\u002Fgoogle-deepmind\u002Fscience-skills)\n(Apache-2.0)._\n",{"name":5,"description":7},"main","e71245365df78e06cd166602744bf2491f79910e","research-skills-hub\u002Fopen-paper-skills\u002Fpaper-search",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 paper-search",{"tool":39,"command":40},"bunx","bunx skills add pengqianhan\u002FAI-Human-Research-OS --skill paper-search",{"tool":42,"command":43},"pnpm","pnpm dlx skills add pengqianhan\u002FAI-Human-Research-OS --skill paper-search","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\u002Fpaper-search\u002FSKILL.md",true,[47,53],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"scripts\u002Fpaper_search.py","SCRIPT","脚本",48374,"https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fopen-paper-skills\u002Fpaper-search\u002Fscripts\u002Fpaper_search.py",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"scripts\u002Ftest_paper_search.py",30280,"https:\u002F\u002Fgithub.com\u002Fpengqianhan\u002FAI-Human-Research-OS\u002Fblob\u002Fmain\u002Fresearch-skills-hub\u002Fopen-paper-skills\u002Fpaper-search\u002Fscripts\u002Ftest_paper_search.py",[58,65,72,79,86,93],{"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},"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":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":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"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":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"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":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"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":87,"name":88,"nameZh":6,"description":89,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":90,"categories":91,"keywords":92,"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":94,"name":95,"nameZh":6,"description":96,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":97,"categories":98,"keywords":99,"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",[],[]]