[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7C3R7CZ6P570ZKDPKYCT7JS":3,"skill-same:01M7C3R7CZ6P570ZKDPKYCT7JS":45},{"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":6,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":30,"install":31,"resources":44},"01M7C3R7CZ6P570ZKDPKYCT7JS","skilltta-task-guidance",null,"Task-specific operational guidance synthesized at test time from retrieved training examples.","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\u002Fskilltta-synthesis-opus47-tatumass-failed15-20260923\u002Ftasks\u002Freserves-at-risk-calc\u002Frepaired_run\u002Finputs\u002Fskills\u002Fskilltta-task-guidance\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: skilltta-task-guidance\ndescription: Task-specific operational guidance synthesized at test time from retrieved training examples.\n---\n\n# SKILL.md\n\n## When to use\nUse this skill for multi-step Excel workbook tasks that require:\n- Downloading an external data source (e.g., an IMF\u002Fcommodity workbook) and extracting a specific series into a named sheet.\n- Populating designated cells across multiple sheets with **Excel formulas** (not precomputed Python values), because the instruction restricts computation to Excel.\n- Performing financial-style calculations: log returns, rolling volatilities (annualized), cross-sheet lookups (INDEX\u002FMATCH or XLOOKUP), and a final risk metric such as Reserves-at-Risk (RaR).\n- Saving the final workbook to an exact output path.\n\nBefore acting, the agent should gather:\n- The exact structure of the target workbook: sheet names, row\u002Fcolumn layout of each \"Answer\" block, and any HINTs (e.g., \"multiply log returns by 100\").\n- The exact URL\u002Fsource and which series\u002Funits to extract (e.g., \"US$ per troy ounce\").\n- Which rows correspond to which sub-step, and the expected formula semantics (latest value vs. average vs. annualized).\n- Whether openpyxl (or similar) is used to write formulas as strings while preserving existing sheet content and formatting.\n\n## Possible Failure Modes\n- Hardcoding numeric results into cells instead of writing Excel formulas — violates \"only use Excel for formula, computation, etc.\"\n- Downloading the wrong sheet\u002Fseries from the source workbook (commodity files often have many series; pick gold in USD\u002Ftroy oz, monthly frequency).\n- Misaligning the monthly time series (e.g., off-by-one when computing log return `LN(P_t\u002FP_{t-1})`, or forgetting to multiply by 100 per HINT).\n- Volatility mistakes:\n  - Using population vs. sample stdev inconsistently.\n  - Forgetting annualization factor (`*SQRT(12)` for monthly data).\n  - Using the wrong rolling window (3 vs. 12) or not using the *latest* window for the Answer sheet.\n- STEP 2 errors:\n  - Missing countries that appear only in the Volume sheet with 2025 data — must be added and converted to Value using Jan–Sep 2025 average gold price.\n  - Using annual or wrong-period average instead of Jan–Sep average as the 2025 price proxy.\n- STEP 3 errors:\n  - Using VLOOKUP or hardcoded joins instead of INDEX+MATCH \u002F XLOOKUP as instructed.\n  - Failing to delete countries lacking 2025 Total Reserves data before computing RaR.\n  - RaR formula misapplied (typical form: `Gold Value × Volatility × Total Reserves ratio`, or `Gold Value \u002F Total Reserves × Volatility` — follow the sheet's row semantics).\n- Overwriting or destroying existing sheets\u002Fformatting when writing with openpyxl (use `load_workbook`, don't recreate).\n- Saving to the wrong path or wrong extension; the target expects `\u002Froot\u002Foutput\u002Frar_result.xlsx` exactly.\n- Copying concrete numbers, sheet layouts, or country lists from retrieved BigCodeBench examples — those are unrelated pandas exercises, not templates for this task.\n\n## Possible procedures\n1. **Fetch source data**\n   - Download the IMF global commodity Excel from the given page. Inspect sheet names; locate the monthly gold price series in USD per troy ounce.\n   - Copy the (Date, Price) columns into the workbook's \"Gold price\" sheet, starting where the template expects.\n\n2. **Compute return\u002Fvolatility formulas (in Excel)**\n   - Column C (monthly log return): `=LN(B3\u002FB2)*100` from the second data row onward.\n   - Column D (3-month volatility): rolling stdev of last 3 log returns, annualized: `=STDEV(C_{t-2}:C_t)*SQRT(12)`.\n   - Column E (12-month volatility): `=STDEV(C_{t-11}:C_t)*SQRT(12)`.\n   - Decide STDEV vs STDEV.S consistently; document choice.\n\n3. **Fill Answer STEP 1 (rows 3–6)**\n   - Use references to the *latest* available row in \"Gold price\" for 3-mo and 12-mo volatility.\n   - For the \"3-month annualized\" cell, derive from the 3-month volatility cell (already annualized if using SQRT(12); otherwise apply the annualization here).\n\n4. **STEP 2 (rows 11–13)**\n   - From \"Value\" sheet: list every country with a non-empty 2025 value; write country names and values via cell references.\n   - Cross-check \"Volume\" sheet: any country with 2025 volume data but absent from \"Value\" must be appended; compute Value = Volume × (Jan–Sep 2025 average gold price, referenced via `AVERAGE` over the relevant Jan–Sep cells in \"Gold price\").\n   - Row 13 \"Gold price exposure\": sum or per-country exposure — read the row label to confirm; typically total gold reserves value.\n\n5. **STEP 3 (rows 20–24)**\n   - Copy country list, values, and volatility from STEP 2 by cell reference.\n   - Row 23 Total Reserves 2025: use `=INDEX(TotalReserves!range, MATCH(country, TotalReserves!keys, 0))` or `=XLOOKUP(country, keys, values2025, \"\")`.\n   - Remove (delete row \u002F blank out) any country whose lookup returns no 2025 data before computing RaR.\n   - Row 24 RaR: apply the formula implied by the sheet header (commonly `Gold Value × Volatility \u002F Total Reserves` — verify against the template's row labels).\n\n6. **Persist**\n   - Open the template with `openpyxl.load_workbook`, write formulas as strings (starting with `=`), save to `\u002Froot\u002Foutput\u002Frar_result.xlsx`. Ensure `\u002Froot\u002Foutput` exists.\n\nDecision points \u002F recovery checks:\n- If the IMF workbook layout changed, search sheets for the string \"Gold\" and unit \"troy\" to locate the correct column.\n- If volatility cells show `#DIV\u002F0!` or `#NUM!`, verify enough prior rows exist for the rolling window.\n- If XLOOKUP is unavailable in the Excel engine used, fall back to INDEX+MATCH.\n- If unsure whether a country belongs in STEP 2's \"Volume-only\" additions, verify it truly has 2025 data and is truly absent from Value (case\u002Fwhitespace-normalize names).\n\n## Verification Checklist\n- [ ] `\u002Froot\u002Foutput\u002Frar_result.xlsx` exists and opens cleanly.\n- [ ] \"Gold price\" sheet contains monthly gold USD\u002Ftroy oz data with columns C (log return ×100), D (3-mo annualized vol), E (12-mo annualized vol), all as **Excel formulas**, not literal numbers.\n- [ ] STEP 1 answer cells (rows 3–6) reference the latest row of \"Gold price\" and include the correctly annualized 3-month figure.\n- [ ] STEP 2 (rows 11–12) includes every 2025 country from \"Value\" plus any 2025-only countries from \"Volume\", with the latter's values computed as `Volume × AVERAGE(Jan–Sep 2025 gold price)` via formula.\n- [ ] Row 13 gold price exposure is a formula, consistent with the row's intent.\n- [ ] STEP 3 (rows 20–22) mirrors STEP 2 via cell references; row 23 uses INDEX+MATCH or XLOOKUP against a Total Reserves sheet; countries without 2025 total reserve data are removed.\n- [ ] Row 24 RaR formula is present and evaluates to a finite number for every remaining country.\n- [ ] No cells contain hardcoded results where a formula is required; no computation was done outside Excel and pasted as values.\n- [ ] Existing sheets, headers, and formatting from the template are preserved; only intended cells were modified.\n- [ ] Nothing was copied from unrelated retrieved examples (they are generic pandas tasks, not templates).\n- [ ] Recovery: if any formula shows an error (`#N\u002FA`, `#DIV\u002F0!`, `#REF!`), trace the referenced range and fix before finalizing.\n","# SKILL.md\n\n## When to use\nUse this skill for multi-step Excel workbook tasks that require:\n- Downloading an external data source (e.g., an IMF\u002Fcommodity workbook) and extracting a specific series into a named sheet.\n- Populating designated cells across multiple sheets with **Excel formulas** (not precomputed Python values), because the instruction restricts computation to Excel.\n- Performing financial-style calculations: log returns, rolling volatilities (annualized), cross-sheet lookups (INDEX\u002FMATCH or XLOOKUP), and a final risk metric such as Reserves-at-Risk (RaR).\n- Saving the final workbook to an exact output path.\n\nBefore acting, the agent should gather:\n- The exact structure of the target workbook: sheet names, row\u002Fcolumn layout of each \"Answer\" block, and any HINTs (e.g., \"multiply log returns by 100\").\n- The exact URL\u002Fsource and which series\u002Funits to extract (e.g., \"US$ per troy ounce\").\n- Which rows correspond to which sub-step, and the expected formula semantics (latest value vs. average vs. annualized).\n- Whether openpyxl (or similar) is used to write formulas as strings while preserving existing sheet content and formatting.\n\n## Possible Failure Modes\n- Hardcoding numeric results into cells instead of writing Excel formulas — violates \"only use Excel for formula, computation, etc.\"\n- Downloading the wrong sheet\u002Fseries from the source workbook (commodity files often have many series; pick gold in USD\u002Ftroy oz, monthly frequency).\n- Misaligning the monthly time series (e.g., off-by-one when computing log return `LN(P_t\u002FP_{t-1})`, or forgetting to multiply by 100 per HINT).\n- Volatility mistakes:\n  - Using population vs. sample stdev inconsistently.\n  - Forgetting annualization factor (`*SQRT(12)` for monthly data).\n  - Using the wrong rolling window (3 vs. 12) or not using the *latest* window for the Answer sheet.\n- STEP 2 errors:\n  - Missing countries that appear only in the Volume sheet with 2025 data — must be added and converted to Value using Jan–Sep 2025 average gold price.\n  - Using annual or wrong-period average instead of Jan–Sep average as the 2025 price proxy.\n- STEP 3 errors:\n  - Using VLOOKUP or hardcoded joins instead of INDEX+MATCH \u002F XLOOKUP as instructed.\n  - Failing to delete countries lacking 2025 Total Reserves data before computing RaR.\n  - RaR formula misapplied (typical form: `Gold Value × Volatility × Total Reserves ratio`, or `Gold Value \u002F Total Reserves × Volatility` — follow the sheet's row semantics).\n- Overwriting or destroying existing sheets\u002Fformatting when writing with openpyxl (use `load_workbook`, don't recreate).\n- Saving to the wrong path or wrong extension; the target expects `\u002Froot\u002Foutput\u002Frar_result.xlsx` exactly.\n- Copying concrete numbers, sheet layouts, or country lists from retrieved BigCodeBench examples — those are unrelated pandas exercises, not templates for this task.\n\n## Possible procedures\n1. **Fetch source data**\n   - Download the IMF global commodity Excel from the given page. Inspect sheet names; locate the monthly gold price series in USD per troy ounce.\n   - Copy the (Date, Price) columns into the workbook's \"Gold price\" sheet, starting where the template expects.\n\n2. **Compute return\u002Fvolatility formulas (in Excel)**\n   - Column C (monthly log return): `=LN(B3\u002FB2)*100` from the second data row onward.\n   - Column D (3-month volatility): rolling stdev of last 3 log returns, annualized: `=STDEV(C_{t-2}:C_t)*SQRT(12)`.\n   - Column E (12-month volatility): `=STDEV(C_{t-11}:C_t)*SQRT(12)`.\n   - Decide STDEV vs STDEV.S consistently; document choice.\n\n3. **Fill Answer STEP 1 (rows 3–6)**\n   - Use references to the *latest* available row in \"Gold price\" for 3-mo and 12-mo volatility.\n   - For the \"3-month annualized\" cell, derive from the 3-month volatility cell (already annualized if using SQRT(12); otherwise apply the annualization here).\n\n4. **STEP 2 (rows 11–13)**\n   - From \"Value\" sheet: list every country with a non-empty 2025 value; write country names and values via cell references.\n   - Cross-check \"Volume\" sheet: any country with 2025 volume data but absent from \"Value\" must be appended; compute Value = Volume × (Jan–Sep 2025 average gold price, referenced via `AVERAGE` over the relevant Jan–Sep cells in \"Gold price\").\n   - Row 13 \"Gold price exposure\": sum or per-country exposure — read the row label to confirm; typically total gold reserves value.\n\n5. **STEP 3 (rows 20–24)**\n   - Copy country list, values, and volatility from STEP 2 by cell reference.\n   - Row 23 Total Reserves 2025: use `=INDEX(TotalReserves!range, MATCH(country, TotalReserves!keys, 0))` or `=XLOOKUP(country, keys, values2025, \"\")`.\n   - Remove (delete row \u002F blank out) any country whose lookup returns no 2025 data before computing RaR.\n   - Row 24 RaR: apply the formula implied by the sheet header (commonly `Gold Value × Volatility \u002F Total Reserves` — verify against the template's row labels).\n\n6. **Persist**\n   - Open the template with `openpyxl.load_workbook`, write formulas as strings (starting with `=`), save to `\u002Froot\u002Foutput\u002Frar_result.xlsx`. Ensure `\u002Froot\u002Foutput` exists.\n\nDecision points \u002F recovery checks:\n- If the IMF workbook layout changed, search sheets for the string \"Gold\" and unit \"troy\" to locate the correct column.\n- If volatility cells show `#DIV\u002F0!` or `#NUM!`, verify enough prior rows exist for the rolling window.\n- If XLOOKUP is unavailable in the Excel engine used, fall back to INDEX+MATCH.\n- If unsure whether a country belongs in STEP 2's \"Volume-only\" additions, verify it truly has 2025 data and is truly absent from Value (case\u002Fwhitespace-normalize names).\n\n## Verification Checklist\n- [ ] `\u002Froot\u002Foutput\u002Frar_result.xlsx` exists and opens cleanly.\n- [ ] \"Gold price\" sheet contains monthly gold USD\u002Ftroy oz data with columns C (log return ×100), D (3-mo annualized vol), E (12-mo annualized vol), all as **Excel formulas**, not literal numbers.\n- [ ] STEP 1 answer cells (rows 3–6) reference the latest row of \"Gold price\" and include the correctly annualized 3-month figure.\n- [ ] STEP 2 (rows 11–12) includes every 2025 country from \"Value\" plus any 2025-only countries from \"Volume\", with the latter's values computed as `Volume × AVERAGE(Jan–Sep 2025 gold price)` via formula.\n- [ ] Row 13 gold price exposure is a formula, consistent with the row's intent.\n- [ ] STEP 3 (rows 20–22) mirrors STEP 2 via cell references; row 23 uses INDEX+MATCH or XLOOKUP against a Total Reserves sheet; countries without 2025 total reserve data are removed.\n- [ ] Row 24 RaR formula is present and evaluates to a finite number for every remaining country.\n- [ ] No cells contain hardcoded results where a formula is required; no computation was done outside Excel and pasted as values.\n- [ ] Existing sheets, headers, and formatting from the template are preserved; only intended cells were modified.\n- [ ] Nothing was copied from unrelated retrieved examples (they are generic pandas tasks, not templates).\n- [ ] Recovery: if any formula shows an error (`#N\u002FA`, `#DIV\u002F0!`, `#REF!`), trace the referenced range and fix before finalizing.\n",{"name":5,"description":7},"main","cdcb6d230f3b15c3c33f8a2bdd16439f43602cbe","evaluation\u002Fresults\u002Fskilltta-synthesis-opus47-tatumass-failed15-20260923\u002Ftasks\u002Freserves-at-risk-calc\u002Frepaired_run\u002Finputs\u002Fskills\u002Fskilltta-task-guidance",1,"https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark","https:\u002F\u002Fgithub.com\u002Flinyouxue","2026-11-14 20:28:35",{"source":8,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add linyouxue\u002Fskill-repair-benchmark --skill skilltta-task-guidance",{"tool":37,"command":38},"bunx","bunx skills add linyouxue\u002Fskill-repair-benchmark --skill skilltta-task-guidance",{"tool":40,"command":41},"pnpm","pnpm dlx skills add linyouxue\u002Fskill-repair-benchmark --skill skilltta-task-guidance","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fevaluation\u002Fresults\u002Fskilltta-synthesis-opus47-tatumass-failed15-20260923\u002Ftasks\u002Freserves-at-risk-calc\u002Frepaired_run\u002Finputs\u002Fskills\u002Fskilltta-task-guidance\u002FSKILL.md",true,[],[46,53,60,67,72,77],{"skillNo":47,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":50,"categories":51,"keywords":52,"publishedAt":19},"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":54,"name":55,"nameZh":6,"description":56,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":57,"categories":58,"keywords":59,"publishedAt":19},"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":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":19},"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":68,"name":48,"nameZh":6,"description":49,"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},"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":73,"name":55,"nameZh":6,"description":56,"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},"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":78,"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":79,"categories":80,"keywords":81,"publishedAt":19},"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",[],[]]