[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7C245B3B9SB00H8FJ76JBPR":3,"skill-same:01M7C245B3B9SB00H8FJ76JBPR":53},{"skillNo":4,"name":5,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":20,"frontMatter":21,"refName":22,"commitSha":23,"skillDirectory":24,"forksCount":25,"repositoryHtmlUrl":26,"repositoryHomepageUrl":6,"ownerHtmlUrl":27,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":28,"install":29,"resources":42},"01M7C245B3B9SB00H8FJ76JBPR","fuzzy-name-search",null,"linyouxue\u002Fskill-repair-benchmark","linyouxue","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F216151464?v=4","01M7BJEVCXKD74KH6FG9CBVSQB","林宥学","01M7BJEVE02NP217V93BKZFC0J",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fevaluation\u002Fresults\u002FSkillrevise\u002Ftasks\u002Fsec-financial-report\u002Frepaired_run\u002Finputs\u002Fskills\u002Ffuzzy-name-search\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"# 13F Filer and Holding Analysis (Discovery-First)\n\n## Purpose\nProvide a reusable, environment-agnostic workflow to identify 13F filers and issuers from quarterly datasets (even with noisy names) and compute common downstream metrics (AUM, holding counts, quarter-to-quarter deltas, and holder rankings) with explicit discovery and schema validation checkpoints.\n\n## When to Use\nUse when a task references SEC 13F quarterly data (one or more quarter folders) and requires any combination of: (a) finding the correct filer from an approximate manager name, (b) mapping issuer name to CUSIP, or (c) computing metrics from 13F tables (AUM from summary pages, counts of holdings, VALUE-based deltas across quarters, or ranking filers by a position). Do not use if the task is unrelated to 13F data or already provides exact identifiers and no computations are needed.\n\n## Procedure\n- 1) Discover data roots and available tables (no hard-coded paths).\n- If the user provides quarter directories, use them; otherwise discover likely roots by listing the working directory and common mounts (for example, the current dir and immediate subdirs).\n- For each quarter directory, search for the key tables by pattern rather than fixed filenames:\n- COVERPAGE (filer identity fields)\n- SUMMARYPAGE (AUM\u002Fsummary fields)\n- INFOTABLE (holdings: issuer\u002FCUSIP\u002FVALUE\u002FSHARES)\n- Execution anchor (required): print the resolved file paths found for each table and print the header row (column names) for each file before proceeding.\n- Decision: if any required table for the requested computation is missing, stop and request the missing location or switch to a computation that only uses available tables.\n- 2) Load tables with schema validation and minimal normalization.\n- Load using a general-purpose reader that matches the discovered format (TSV\u002FCSV); do not assume delimiter until verified by inspecting the file (header and a couple lines).\n- Validate required columns exist before using them:\n- Filer identification typically needs an accession-like identifier plus name\u002Faddress fields from COVERPAGE.\n- AUM typically needs a numeric total from SUMMARYPAGE.\n- Holdings computations need at least: accession identifier, CUSIP (or issuer identifier), and VALUE from INFOTABLE.\n- Parse numeric fields defensively (strip commas\u002Fwhitespace); fail fast if parsing yields mostly missing\u002FNaN.\n- Execution anchor (required): print row counts and a small sanity summary (for example, distinct accession count; min\u002Fmax of VALUE if present). Branch\u002Ffail if tables are empty or columns missing.\n- 3) Identify the correct filer (exact match first, then fuzzy with disambiguation).\n- If an exact identifier is provided (accession\u002FCIK\u002F13F file number), filter by it and verify exactly one (or a clearly best) match.\n- Otherwise fuzzy-match on normalized names:\n- Normalize: lowercase, collapse whitespace, remove punctuation-only differences.\n- Retrieve top K candidates with scores, but do not auto-select if the top scores are close.\n- Disambiguate using additional constraints when provided (city\u002Fstate, filing number, founder facts, etc.) by checking those fields in COVERPAGE rather than guessing.\n- Decision: if ambiguity remains (top candidates within a small score band or conflicting constraints), stop and present the candidate list and what extra info is needed to choose.\n- 4) Compute requested 13F metrics with cross-checks.\n- AUM extraction: read the SUMMARYPAGE row(s) for the chosen accession and extract the declared total; record the units as stated by the dataset (do not assume dollars vs thousands without evidence).\n- Holding count: in INFOTABLE filtered to the accession, count distinct holdings using a stable key (CUSIP preferred); define whether duplicates are possible and deduplicate explicitly.\n- Quarter-to-quarter deltas: for two quarters, resolve the same filer across quarters using identifiers (preferred) or validated name+address matching; compute per-CUSIP VALUE differences after aligning units and types.\n- Rank holders of a target CUSIP: for a given quarter, filter INFOTABLE by CUSIP and aggregate VALUE by accession\u002Ffiler, then sort descending; verify that the CUSIP exists in the table before ranking.\n- Final checkpoint: re-run a minimal verification on the computed outputs (non-empty results, expected keys present, and totals plausible relative to raw VALUE ranges). If any check fails, return to step 1\u002F2 to re-verify paths, delimiters, and columns.\n\n## Constraints \u002F Pitfalls\n- Never hard-code repo-specific scripts, flags, quarter labels, or absolute paths. Always discover files and verify headers before running any parsing or computation.\n- Do not assume units for VALUE or AUM. Only label\u002Fconvert units when the dataset explicitly indicates them; otherwise keep values in raw reported units and state that explicitly in downstream computations.",{},"main","cdcb6d230f3b15c3c33f8a2bdd16439f43602cbe","evaluation\u002Fresults\u002FSkillrevise\u002Ftasks\u002Fsec-financial-report\u002Frepaired_run\u002Finputs\u002Fskills\u002Ffuzzy-name-search",1,"https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark","https:\u002F\u002Fgithub.com\u002Flinyouxue","2026-11-14 20:28:35",{"source":7,"skillName":5,"commands":30,"prompt":40,"zipAvailable":41},[31,34,37],{"tool":32,"command":33},"npx","npx skills add linyouxue\u002Fskill-repair-benchmark --skill fuzzy-name-search",{"tool":35,"command":36},"bunx","bunx skills add linyouxue\u002Fskill-repair-benchmark --skill fuzzy-name-search",{"tool":38,"command":39},"pnpm","pnpm dlx skills add linyouxue\u002Fskill-repair-benchmark --skill fuzzy-name-search","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fevaluation\u002Fresults\u002FSkillrevise\u002Ftasks\u002Fsec-financial-report\u002Frepaired_run\u002Finputs\u002Fskills\u002Ffuzzy-name-search\u002FSKILL.md",true,[43,49],{"path":44,"resourceType":45,"resourceTypeDesc":46,"sizeBytes":47,"sourceHtmlUrl":48},"scripts\u002Fsearch_fund.py","SCRIPT","脚本",3604,"https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fevaluation\u002Fresults\u002FSkillrevise\u002Ftasks\u002Fsec-financial-report\u002Frepaired_run\u002Finputs\u002Fskills\u002Ffuzzy-name-search\u002Fscripts\u002Fsearch_fund.py",{"path":50,"resourceType":45,"resourceTypeDesc":46,"sizeBytes":51,"sourceHtmlUrl":52},"scripts\u002Fsearch_stock_cusip.py",2359,"https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fevaluation\u002Fresults\u002FSkillrevise\u002Ftasks\u002Fsec-financial-report\u002Frepaired_run\u002Finputs\u002Fskills\u002Ffuzzy-name-search\u002Fscripts\u002Fsearch_stock_cusip.py",[54,61,68,75,80,85],{"skillNo":55,"name":56,"nameZh":6,"description":57,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":60,"publishedAt":18},"01M7E1HGVRY73NBJ022CJG13X6","video-frame-extraction","Extract frames from video files and save them as images using OpenCV","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_03\u002Finputs\u002Fskills\u002Fvideo-frame-extraction\u002FSKILL.md",[],[],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":18},"01M7E1HAG5R7CJ9C8NYJS771MB","openai-vision","Analyze images and multi-frame sequences using OpenAI GPT vision models","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_03\u002Finputs\u002Fskills\u002Fopenai-vision\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":18},"01M7E1H7DADKB6QBWKC2KQ63SZ","image-ocr","Extract text content from images using Tesseract OCR via Python, including structured field extraction from receipts and forms","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_03\u002Finputs\u002Fskills\u002Fimage-ocr\u002FSKILL.md",[],[],{"skillNo":76,"name":56,"nameZh":6,"description":57,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":18},"01M7E1H0WBNT43TENGQY7AVZHD","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_02\u002Finputs\u002Fskills\u002Fvideo-frame-extraction\u002FSKILL.md",[],[],{"skillNo":81,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":18},"01M7E1GTC7D4N4Q2EF03875J9E","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_02\u002Finputs\u002Fskills\u002Fopenai-vision\u002FSKILL.md",[],[],{"skillNo":86,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":18},"01M7E1GQ92CWFMGH53KSSE4WER","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_02\u002Finputs\u002Fskills\u002Fimage-ocr\u002FSKILL.md",[],[]]