[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6KFVQHWY79E53PAGPSTCB8V":3,"skill-same:01M6KFVQHWY79E53PAGPSTCB8V":59},{"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":26,"linkOnly":27,"markdownContent":28,"markdownBody":29,"frontMatter":30,"refName":31,"commitSha":32,"skillDirectory":33,"forksCount":14,"repositoryHtmlUrl":34,"repositoryHomepageUrl":6,"ownerHtmlUrl":35,"ownerBlogUrl":36,"ownerDescription":37,"updatedAt":38,"install":39,"resources":52},"01M6KFVQHWY79E53PAGPSTCB8V","prioritize-reference-files",null,"Ensures agents read and use provided reference files before searching or fabricating data","xyliu-cs\u002FOpenSpace-GDPval","xyliu-cs","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F89236536?v=4","01M6KF8Z1V4TXS86CKSH48BY8T","xy","01M6KF8Z67443W0SHXM85JEF0B",0,"MIT","https:\u002F\u002Fgithub.com\u002Fxyliu-cs\u002FOpenSpace-GDPval\u002Fblob\u002Fmain\u002Fgdpval_bench\u002Fskills\u002Fprioritize-reference-files\u002FSKILL.md",[],[19,20,21,22,23,24,25],"prioritize","reference","files","prioritize reference files","data","read","use","2026-11-10 10:12:30",false,"---\nname: prioritize-reference-files\ndescription: Ensures agents read and use provided reference files before searching or fabricating data\n---\n\n# Prioritize Reference Files\n\n## Purpose\n\nThis skill ensures that when reference files are provided in task context, you MUST read and extract data from them FIRST before attempting web searches or generating synthetic data. Ignoring available structured data leads to fabricated outputs and incorrect results.\n\n## Core Principle\n\n**Reference files in context > Web search > Data fabrication (never)**\n\n## Workflow\n\n### Step 1: Scan Task Context for Reference Files\n\nBefore taking any action, identify all files provided in the task context:\n- Look for file attachments, uploads, or references in the task description\n- Common formats: `.xlsx`, `.csv`, `.json`, `.pdf`, `.docx`, `.txt`\n- Check for phrases like \"attached\", \"provided\", \"reference file\", \"see file\"\n\n### Step 2: Read Reference Files First\n\nUse the appropriate tool to read each reference file:\n\n```python\n# For Excel files\nread_file(filetype=\"xlsx\", file_path=\"path\u002Fto\u002Ffile.xlsx\")\n\n# For CSV files  \nread_file(filetype=\"csv\", file_path=\"path\u002Fto\u002Ffile.csv\")\n\n# For PDF files\nread_file(filetype=\"pdf\", file_path=\"path\u002Fto\u002Ffile.pdf\")\n\n# For JSON files\nread_file(filetype=\"json\", file_path=\"path\u002Fto\u002Ffile.json\")\n\n# For text files\nread_file(filetype=\"txt\", file_path=\"path\u002Fto\u002Ffile.txt\")\n```\n\n**If read_file fails on .docx files** (returns error, empty content, or 'unknown error'):\n\n**Fallback Approach 1: Direct zipfile\u002FXML extraction via run_shell**\n```bash\n# .docx files are ZIP archives containing XML; extract document.xml directly\nunzip -p path\u002Fto\u002Ffile.docx word\u002Fdocument.xml | grep -oP '(?\u003C=\u003Cw:t>)[^\u003C]+' | tr '\\n' ' '\n```\nOr for more complete extraction:\n```bash\nmkdir -p \u002Ftmp\u002Fdocx_extract && cd \u002Ftmp\u002Fdocx_extract && unzip path\u002Fto\u002Ffile.docx && cat word\u002Fdocument.xml\n```\n\n**Fallback Approach 2: Use shell_agent for complex extraction**\nIf direct extraction fails, delegate to shell_agent:\n```\nshell_agent(task=\"Extract text content from path\u002Fto\u002Ffile.docx using zipfile and XML parsing\")\n```\nThe agent will attempt multiple extraction methods and report results.\n\n**Fallback Approach 3: Verify extraction success before proceeding**\nAfter any extraction method, confirm content was retrieved:\n- Check that output is non-empty and contains expected document structure\n- Look for familiar text from the document (headings, known phrases)\n- If extraction appears incomplete, try an alternative method\n- Document which method succeeded: \"Content extracted using [method] after read_file failed\"\n\n**Important**: Never proceed to data fabrication if reference files exist but read_file fails. Always attempt at least one fallback extraction method first.\n\n### Step 3: Extract and Validate Data\n\nAfter reading:\n1. Parse the content to understand the data structure\n2. Identify relevant fields\u002Fcolumns for your task\n3. Verify the data is complete and usable\n4. Note any limitations or missing information\n\n### Step 4: Use Reference Data as Primary Source\n\n- Base all outputs on the reference file data\n- Only supplement with web search if reference data is incomplete\n- NEVER fabricate data when reference files exist\n- If reference data is insufficient, explicitly state what's missing\n\n### Step 5: Document Data Source\n\nIn your outputs, acknowledge the source:\n- \"Data sourced from [filename]\"\n- \"Based on provided reference file: [filename]\"\n- This confirms you used the actual provided data\n\n## Anti-Patterns to Avoid\n\n❌ **Ignoring reference files** and searching the web instead\n❌ **Fabricating data** when structured data is available\n❌ **Assuming file contents** without reading them\n❌ **Using outdated web data** when current reference files exist\n❌ **Giving up after read_file fails** without trying fallback extraction methods\n\n## Example\n\n**Task Context**: \"Create a property listings report. See Massabama_active_listings.xlsx for current data.\"\n\n**Correct Approach**:\n```\n1. Read Massabama_active_listings.xlsx first\n2. Extract property addresses, prices, specifications\n3. Generate report using actual listing data\n4. Note: \"Data sourced from Massabama_active_listings.xlsx\"\n```\n\n**Incorrect Approach**:\n```\n1. Search web for \"Massabama property listings\"\n2. Fabricate property data from search results\n3. Create report with unverified\u002Fgenerated data\n```\n\n## Verification Checklist\n\nBefore completing any task with reference files:\n\n- [ ] I have identified all reference files in the task context\n- [ ] I have read each reference file using appropriate tools\n- [ ] I have extracted relevant data from the files\n- [ ] My output is based on the reference file data\n- [ ] I have documented the data source in my output\n\n## When Web Search is Appropriate\n\nOnly search the web when:\n- Reference files don't exist for the required data\n- Reference files are incomplete AND supplemental info is needed\n- You need to verify or update time-sensitive information\n- You explicitly state what the reference files lack\n\n---\n","# Prioritize Reference Files\n\n## Purpose\n\nThis skill ensures that when reference files are provided in task context, you MUST read and extract data from them FIRST before attempting web searches or generating synthetic data. Ignoring available structured data leads to fabricated outputs and incorrect results.\n\n## Core Principle\n\n**Reference files in context > Web search > Data fabrication (never)**\n\n## Workflow\n\n### Step 1: Scan Task Context for Reference Files\n\nBefore taking any action, identify all files provided in the task context:\n- Look for file attachments, uploads, or references in the task description\n- Common formats: `.xlsx`, `.csv`, `.json`, `.pdf`, `.docx`, `.txt`\n- Check for phrases like \"attached\", \"provided\", \"reference file\", \"see file\"\n\n### Step 2: Read Reference Files First\n\nUse the appropriate tool to read each reference file:\n\n```python\n# For Excel files\nread_file(filetype=\"xlsx\", file_path=\"path\u002Fto\u002Ffile.xlsx\")\n\n# For CSV files  \nread_file(filetype=\"csv\", file_path=\"path\u002Fto\u002Ffile.csv\")\n\n# For PDF files\nread_file(filetype=\"pdf\", file_path=\"path\u002Fto\u002Ffile.pdf\")\n\n# For JSON files\nread_file(filetype=\"json\", file_path=\"path\u002Fto\u002Ffile.json\")\n\n# For text files\nread_file(filetype=\"txt\", file_path=\"path\u002Fto\u002Ffile.txt\")\n```\n\n**If read_file fails on .docx files** (returns error, empty content, or 'unknown error'):\n\n**Fallback Approach 1: Direct zipfile\u002FXML extraction via run_shell**\n```bash\n# .docx files are ZIP archives containing XML; extract document.xml directly\nunzip -p path\u002Fto\u002Ffile.docx word\u002Fdocument.xml | grep -oP '(?\u003C=\u003Cw:t>)[^\u003C]+' | tr '\\n' ' '\n```\nOr for more complete extraction:\n```bash\nmkdir -p \u002Ftmp\u002Fdocx_extract && cd \u002Ftmp\u002Fdocx_extract && unzip path\u002Fto\u002Ffile.docx && cat word\u002Fdocument.xml\n```\n\n**Fallback Approach 2: Use shell_agent for complex extraction**\nIf direct extraction fails, delegate to shell_agent:\n```\nshell_agent(task=\"Extract text content from path\u002Fto\u002Ffile.docx using zipfile and XML parsing\")\n```\nThe agent will attempt multiple extraction methods and report results.\n\n**Fallback Approach 3: Verify extraction success before proceeding**\nAfter any extraction method, confirm content was retrieved:\n- Check that output is non-empty and contains expected document structure\n- Look for familiar text from the document (headings, known phrases)\n- If extraction appears incomplete, try an alternative method\n- Document which method succeeded: \"Content extracted using [method] after read_file failed\"\n\n**Important**: Never proceed to data fabrication if reference files exist but read_file fails. Always attempt at least one fallback extraction method first.\n\n### Step 3: Extract and Validate Data\n\nAfter reading:\n1. Parse the content to understand the data structure\n2. Identify relevant fields\u002Fcolumns for your task\n3. Verify the data is complete and usable\n4. Note any limitations or missing information\n\n### Step 4: Use Reference Data as Primary Source\n\n- Base all outputs on the reference file data\n- Only supplement with web search if reference data is incomplete\n- NEVER fabricate data when reference files exist\n- If reference data is insufficient, explicitly state what's missing\n\n### Step 5: Document Data Source\n\nIn your outputs, acknowledge the source:\n- \"Data sourced from [filename]\"\n- \"Based on provided reference file: [filename]\"\n- This confirms you used the actual provided data\n\n## Anti-Patterns to Avoid\n\n❌ **Ignoring reference files** and searching the web instead\n❌ **Fabricating data** when structured data is available\n❌ **Assuming file contents** without reading them\n❌ **Using outdated web data** when current reference files exist\n❌ **Giving up after read_file fails** without trying fallback extraction methods\n\n## Example\n\n**Task Context**: \"Create a property listings report. See Massabama_active_listings.xlsx for current data.\"\n\n**Correct Approach**:\n```\n1. Read Massabama_active_listings.xlsx first\n2. Extract property addresses, prices, specifications\n3. Generate report using actual listing data\n4. Note: \"Data sourced from Massabama_active_listings.xlsx\"\n```\n\n**Incorrect Approach**:\n```\n1. Search web for \"Massabama property listings\"\n2. Fabricate property data from search results\n3. Create report with unverified\u002Fgenerated data\n```\n\n## Verification Checklist\n\nBefore completing any task with reference files:\n\n- [ ] I have identified all reference files in the task context\n- [ ] I have read each reference file using appropriate tools\n- [ ] I have extracted relevant data from the files\n- [ ] My output is based on the reference file data\n- [ ] I have documented the data source in my output\n\n## When Web Search is Appropriate\n\nOnly search the web when:\n- Reference files don't exist for the required data\n- Reference files are incomplete AND supplemental info is needed\n- You need to verify or update time-sensitive information\n- You explicitly state what the reference files lack\n\n---\n",{"name":5,"description":7},"main","af1b40a42167d09c30b3a63000aa14ace4c1e229","gdpval_bench\u002Fskills\u002Fprioritize-reference-files","https:\u002F\u002Fgithub.com\u002Fxyliu-cs\u002FOpenSpace-GDPval","https:\u002F\u002Fgithub.com\u002Fxyliu-cs","https:\u002F\u002Fxyliu-cs.github.io","PhD Student in CS, CUHK-Shenzhen.","2026-04-06 20:32:15",{"source":8,"skillName":5,"commands":40,"prompt":50,"zipAvailable":51},[41,44,47],{"tool":42,"command":43},"npx","npx skills add xyliu-cs\u002FOpenSpace-GDPval --skill prioritize-reference-files",{"tool":45,"command":46},"bunx","bunx skills add xyliu-cs\u002FOpenSpace-GDPval --skill prioritize-reference-files",{"tool":48,"command":49},"pnpm","pnpm dlx skills add xyliu-cs\u002FOpenSpace-GDPval --skill prioritize-reference-files","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fxyliu-cs\u002FOpenSpace-GDPval\u002Fblob\u002Fmain\u002Fgdpval_bench\u002Fskills\u002Fprioritize-reference-files\u002FSKILL.md",true,[53],{"path":54,"resourceType":55,"resourceTypeDesc":56,"sizeBytes":57,"sourceHtmlUrl":58},".skill_id","OTHER","其他",40,"https:\u002F\u002Fgithub.com\u002Fxyliu-cs\u002FOpenSpace-GDPval\u002Fblob\u002Fmain\u002Fgdpval_bench\u002Fskills\u002Fprioritize-reference-files\u002F.skill_id",[60,73,81,97,122,144],{"skillNo":61,"name":62,"nameZh":6,"description":63,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":64,"categories":65,"keywords":66,"publishedAt":26},"01M6KG8MF1S664SWBDENYRSCK8","zip-export-deliverable","Ensures codebase deliverables are properly exported as ZIP archives through a dedicated finalization step","https:\u002F\u002Fgithub.com\u002Fxyliu-cs\u002FOpenSpace-GDPval\u002Fblob\u002Fmain\u002Fgdpval_bench\u002Fskills\u002Fzip-export-deliverable\u002FSKILL.md",[],[67,68,69,70,71,72],"zip export deliverable","zip","export","deliverables","codebase","as",{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":26},"01M6KG8GAZWHHEK7VNJDW0ME21","zip-deliverable-finalization","Ensure codebase deliverables are properly packaged as ZIP exports in a dedicated final step","https:\u002F\u002Fgithub.com\u002Fxyliu-cs\u002FOpenSpace-GDPval\u002Fblob\u002Fmain\u002Fgdpval_bench\u002Fskills\u002Fzip-deliverable-finalization\u002FSKILL.md",[],[80,68,70,71,72],"zip deliverable finalization",{"skillNo":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":85,"categories":86,"keywords":87,"publishedAt":26},"01M6KG8CMQQ1PYMNTSRZ9VS1MD","zero-iteration-failure-analysis","Identify and handle agent failures with 0 iterations as pre-execution system issues","https:\u002F\u002Fgithub.com\u002Fxyliu-cs\u002FOpenSpace-GDPval\u002Fblob\u002Fmain\u002Fgdpval_bench\u002Fskills\u002Fzero-iteration-failure-analysis\u002FSKILL.md",[],[88,89,90,91,92,93,94,95,72,96],"failure analysis","ZeRO","analysis","iteration","zero iteration failure analysis","with","issues","execution","system",{"skillNo":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":108,"publishedAt":26},"01M6KG86W3PA07FT9W58YHHERM","write-file-fallback-report","Generate professional documents using write_file when primary data sources fail, leveraging embedded domain knowledge instead of external retrieval","https:\u002F\u002Fgithub.com\u002Fxyliu-cs\u002FOpenSpace-GDPval\u002Fblob\u002Fmain\u002Fgdpval_bench\u002Fskills\u002Fwrite-file-fallback-report\u002FSKILL.md",[103],{"categoryNo":104,"categoryName":105,"taxonomyType":106,"taxonomyTypeDesc":107},"01JBSEEDCAT000000000000034","行政与办公人员","PROFESSION","适用职业",[109,110,111,112,113,114,115,116,117,118,119,120,23,121],"fallback","write","file","report","write file fallback report","domain","professional","sources","retrieval","generate","embedded","documents","knowledge",{"skillNo":123,"name":124,"nameZh":6,"description":125,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":126,"categories":127,"keywords":133,"publishedAt":26},"01M6KG82GAEY3GG77CMGCK1C2T","fallback-doc-with-delegation","Generate professional documents after tool failures using shell_agent for complex formats or write_file for simple text\u002Fmarkdown","https:\u002F\u002Fgithub.com\u002Fxyliu-cs\u002FOpenSpace-GDPval\u002Fblob\u002Fmain\u002Fgdpval_bench\u002Fskills\u002Fwrite-file-fallback-report-enhanced\u002FSKILL.md",[128],{"categoryNo":129,"categoryName":130,"taxonomyType":131,"taxonomyTypeDesc":132},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[134,93,109,135,136,137,120,118,111,138,139,140,110,141,115,142,143],"fallback doc with delegation","doc","delegation","shell","markdown","tool","text","simple","formats","for",{"skillNo":145,"name":146,"nameZh":6,"description":147,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":148,"categories":149,"keywords":151,"publishedAt":26},"01M6KG7WT7GX5NYCGEFHW22NV7","context-anchored-fallback-report","Generate documents with write_file when retrieval tools fail, with explicit guardrails against task context drift","https:\u002F\u002Fgithub.com\u002Fxyliu-cs\u002FOpenSpace-GDPval\u002Fblob\u002Fmain\u002Fgdpval_bench\u002Fskills\u002Fwrite-file-fallback-report-enhanced-4bc5c2\u002FSKILL.md",[150],{"categoryNo":104,"categoryName":105,"taxonomyType":106,"taxonomyTypeDesc":107},[109,112,152,153,118,111,154,155,110,156,157,117,93,158,120],"context","context anchored fallback report","Guardrails","Tools","task","drift","context drift"]