[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7C2XJBGGC4C1P9DADRNRYH2":3,"skill-same:01M7C2XJBGGC4C1P9DADRNRYH2":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},"01M7C2XJBGGC4C1P9DADRNRYH2","milp-solver-workflow",null,"Use for formulating, solving, debugging, and validating mixed-integer linear optimization models with open-source solvers, including variable indexing, sparse constraints, linearized costs, solver limits, MIP gaps, incumbent extraction, numerical tolerances, and deterministic output reporting.","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\u002FWML\u002Ftasks\u002Fenergy-unit-commitment\u002Frepaired_skill\u002Fmilp-solver-workflow\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: milp-solver-workflow\ndescription: Use for formulating, solving, debugging, and validating mixed-integer linear optimization models with open-source solvers, including variable indexing, sparse constraints, linearized costs, solver limits, MIP gaps, incumbent extraction, numerical tolerances, and deterministic output reporting.\n---\n\n# MILP Solver Workflow\n\nUse this skill for binary\u002Finteger decisions, linear constraints, and linear or piecewise-linear objectives. It is useful for time-expanded scheduling models with many repeated resource-period constraints.\n\nThis is a workflow and implementation guide, not a complete formulation for any one task.\n\n## Workflow\n\n1. Parse and normalize data into ordered arrays.\n2. Define decision states before coding: status, transitions, continuous quantities, slacks, segments, tiers.\n3. Build a deterministic variable map.\n4. Add constraints family by family: bounds, linking, balance, time coupling, capacity\u002Framp limits, cost logic.\n5. Solve with an available open-source MILP solver.\n6. Extract a candidate solution, rounding binaries only if near integral.\n7. Convert internal variables into the report convention.\n8. Independently validate extracted arrays.\n9. Recompute objective and summaries from extracted arrays.\n10. Write final output only after validation passes.\n\n## Variable Map Pattern\n\nUse helper functions or dictionaries, not scattered index arithmetic.\n\n```python\noffset = {}\nn = 0\n\ndef alloc(name, shape, lb=0.0, ub=float(\"inf\"), integer=False):\n    global n\n    size = int(np.prod(shape))\n    idx = np.arange(n, n + size).reshape(shape)\n    offset[name] = idx\n    n += size\n    return idx\n\nu = alloc(\"commitment\", (G, T), lb=0, ub=1, integer=True)\nstart = alloc(\"startup\", (G, T), lb=0, ub=1, integer=True)\ndispatch = alloc(\"dispatch\", (G, T), lb=0)\nreserve = alloc(\"reserve\", (G, T), lb=0)\n```\n\nKeep variable ownership obvious: type, resource, period, and optional segment\u002Ftier.\n\n## Sparse Constraint Pattern\n\nUse sparse rows for large time-expanded models:\n\n```python\nrows, cols, vals = [], [], []\nlb, ub = [], []\nrow = 0\n\ndef add_row(terms, lo, hi):\n    global row\n    for j, a in terms:\n        if abs(a) > 0:\n            rows.append(row)\n            cols.append(j)\n            vals.append(float(a))\n    lb.append(float(lo))\n    ub.append(float(hi))\n    row += 1\n```\n\nUse equality rows for conservation\u002Flinking and upper\u002Flower bound rows for capacity, reserve, ramping, timing, and logic.\n\n## Sign-Safe Encoding\n\nWrite the intended inequality first, then move variable terms to the left-hand side.\n\n```python\n# Intended: x + y \u003C= cap * u - reduction * start\n# Row form: x + y - cap*u + reduction*start \u003C= 0\nadd_row(\n    [(x, 1.0), (y, 1.0), (u, -cap), (start, reduction)],\n    lo=-INF,\n    hi=0.0,\n)\n```\n\nFor non-obvious rows, test a tiny hand case. Example: set `u=1`, `start=1` and check the remaining capacity equals the intended startup capability; set `u=1`, `start=0` and check normal capacity returns.\n\n## Match Model Rows To Validation\n\nKeep a checklist linking each model constraint family to a validation check:\n\n```text\ntransition linking      -> startup\u002Fshutdown match status\nonline capacity         -> offline zeroes and min\u002Fmax output\njoint reserve capacity  -> production plus reserve fits capability\nramp deliverability     -> reserve can be deployed within ramp-up\nminimum durations       -> starts\u002Fstops imply required status windows\nbalance equations       -> demand\u002Fload\u002Finventory conservation\ncost-curve logic        -> recomputed objective matches report\n```\n\nIf validation checks something not in the model, the solver may produce an invalid report. If the model has a constraint not validated after extraction, conversion bugs can slip through.\n\n## Piecewise-Linear Costs\n\nIdentify the curve convention before modeling:\n\n- total cost at output breakpoints;\n- marginal or incremental segment cost;\n- heat-rate curve;\n- first point as minimum-output or no-load-like cost.\n\nFor segment variables, constrain segment quantities to their widths and sum them to the modeled production quantity. Use slopes only when the data represents total-cost breakpoints or incremental segment costs consistently.\n\n## Open-Source Solver Use\n\nHiGHS through SciPy is a common default when available:\n\n```python\nfrom scipy.optimize import milp, LinearConstraint, Bounds\n\nresult = milp(\n    c=c,\n    integrality=integrality,\n    bounds=Bounds(lb, ub),\n    constraints=constraints,\n    options={\"time_limit\": 600.0, \"mip_rel_gap\": 0.01, \"disp\": False},\n)\n\nif result.x is None:\n    raise RuntimeError(f\"No incumbent: status={result.status}, message={result.message}\")\n```\n\nCapture status, objective, incumbent values, and any reliable gap\u002Fbound. A time-limit status can be useful if a feasible incumbent exists; a failure with no incumbent is not a solution.\n\n## Extraction And Validation\n\nAfter solving:\n\n- check binary variables are close to 0\u002F1 before rounding;\n- convert internal units\u002Fconventions to report units;\n- recompute summaries from arrays;\n- recompute objective from input data and extracted decisions;\n- run validation that reads only input data plus extracted\u002Freport arrays;\n- write `\"pass\"` self-checks only after validation passes.\n\n## Debugging Infeasibility\n\nFirst suspect the model encoding. Common causes:\n\n- mixing total output with output above minimum;\n- applying startup\u002Fshutdown limits to the wrong quantity;\n- using the wrong `t` or `t-1` index;\n- over-constraining initial minimum up\u002Fdown obligations;\n- enforcing post-horizon obligations when the prompt excludes them;\n- treating cost curves as feasibility constraints;\n- choosing bad Big-M values;\n- requiring segment\u002Ftier variables when the trigger did not occur.\n\nDebug in stages: check shapes, relax one family at a time, add diagnostic slack variables, print largest violations, and compare local validation with final requirements.\n\n## Repair LPs And Heuristics\n\nA fixed-commitment repair LP is useful only if it includes every feasibility family judged in the final report. Do not repair only balance and capacity while omitting transition-dependent ramp, reserve, startup, shutdown, or minimum-duration limits.\n\nAlways rerun independent validation after repair.\n\n## Reporting Discipline\n\n- Use deterministic ordering and plain numeric values.\n- Do not include placeholder values.\n- Keep feasibility separate from proof quality\u002FMIP gap.\n- Use `null`\u002Fempty gap when no reliable bound exists, if the schema allows it.\n- Do not trust solver status or self-reported `\"pass\"` strings without validation.\n","# MILP Solver Workflow\n\nUse this skill for binary\u002Finteger decisions, linear constraints, and linear or piecewise-linear objectives. It is useful for time-expanded scheduling models with many repeated resource-period constraints.\n\nThis is a workflow and implementation guide, not a complete formulation for any one task.\n\n## Workflow\n\n1. Parse and normalize data into ordered arrays.\n2. Define decision states before coding: status, transitions, continuous quantities, slacks, segments, tiers.\n3. Build a deterministic variable map.\n4. Add constraints family by family: bounds, linking, balance, time coupling, capacity\u002Framp limits, cost logic.\n5. Solve with an available open-source MILP solver.\n6. Extract a candidate solution, rounding binaries only if near integral.\n7. Convert internal variables into the report convention.\n8. Independently validate extracted arrays.\n9. Recompute objective and summaries from extracted arrays.\n10. Write final output only after validation passes.\n\n## Variable Map Pattern\n\nUse helper functions or dictionaries, not scattered index arithmetic.\n\n```python\noffset = {}\nn = 0\n\ndef alloc(name, shape, lb=0.0, ub=float(\"inf\"), integer=False):\n    global n\n    size = int(np.prod(shape))\n    idx = np.arange(n, n + size).reshape(shape)\n    offset[name] = idx\n    n += size\n    return idx\n\nu = alloc(\"commitment\", (G, T), lb=0, ub=1, integer=True)\nstart = alloc(\"startup\", (G, T), lb=0, ub=1, integer=True)\ndispatch = alloc(\"dispatch\", (G, T), lb=0)\nreserve = alloc(\"reserve\", (G, T), lb=0)\n```\n\nKeep variable ownership obvious: type, resource, period, and optional segment\u002Ftier.\n\n## Sparse Constraint Pattern\n\nUse sparse rows for large time-expanded models:\n\n```python\nrows, cols, vals = [], [], []\nlb, ub = [], []\nrow = 0\n\ndef add_row(terms, lo, hi):\n    global row\n    for j, a in terms:\n        if abs(a) > 0:\n            rows.append(row)\n            cols.append(j)\n            vals.append(float(a))\n    lb.append(float(lo))\n    ub.append(float(hi))\n    row += 1\n```\n\nUse equality rows for conservation\u002Flinking and upper\u002Flower bound rows for capacity, reserve, ramping, timing, and logic.\n\n## Sign-Safe Encoding\n\nWrite the intended inequality first, then move variable terms to the left-hand side.\n\n```python\n# Intended: x + y \u003C= cap * u - reduction * start\n# Row form: x + y - cap*u + reduction*start \u003C= 0\nadd_row(\n    [(x, 1.0), (y, 1.0), (u, -cap), (start, reduction)],\n    lo=-INF,\n    hi=0.0,\n)\n```\n\nFor non-obvious rows, test a tiny hand case. Example: set `u=1`, `start=1` and check the remaining capacity equals the intended startup capability; set `u=1`, `start=0` and check normal capacity returns.\n\n## Match Model Rows To Validation\n\nKeep a checklist linking each model constraint family to a validation check:\n\n```text\ntransition linking      -> startup\u002Fshutdown match status\nonline capacity         -> offline zeroes and min\u002Fmax output\njoint reserve capacity  -> production plus reserve fits capability\nramp deliverability     -> reserve can be deployed within ramp-up\nminimum durations       -> starts\u002Fstops imply required status windows\nbalance equations       -> demand\u002Fload\u002Finventory conservation\ncost-curve logic        -> recomputed objective matches report\n```\n\nIf validation checks something not in the model, the solver may produce an invalid report. If the model has a constraint not validated after extraction, conversion bugs can slip through.\n\n## Piecewise-Linear Costs\n\nIdentify the curve convention before modeling:\n\n- total cost at output breakpoints;\n- marginal or incremental segment cost;\n- heat-rate curve;\n- first point as minimum-output or no-load-like cost.\n\nFor segment variables, constrain segment quantities to their widths and sum them to the modeled production quantity. Use slopes only when the data represents total-cost breakpoints or incremental segment costs consistently.\n\n## Open-Source Solver Use\n\nHiGHS through SciPy is a common default when available:\n\n```python\nfrom scipy.optimize import milp, LinearConstraint, Bounds\n\nresult = milp(\n    c=c,\n    integrality=integrality,\n    bounds=Bounds(lb, ub),\n    constraints=constraints,\n    options={\"time_limit\": 600.0, \"mip_rel_gap\": 0.01, \"disp\": False},\n)\n\nif result.x is None:\n    raise RuntimeError(f\"No incumbent: status={result.status}, message={result.message}\")\n```\n\nCapture status, objective, incumbent values, and any reliable gap\u002Fbound. A time-limit status can be useful if a feasible incumbent exists; a failure with no incumbent is not a solution.\n\n## Extraction And Validation\n\nAfter solving:\n\n- check binary variables are close to 0\u002F1 before rounding;\n- convert internal units\u002Fconventions to report units;\n- recompute summaries from arrays;\n- recompute objective from input data and extracted decisions;\n- run validation that reads only input data plus extracted\u002Freport arrays;\n- write `\"pass\"` self-checks only after validation passes.\n\n## Debugging Infeasibility\n\nFirst suspect the model encoding. Common causes:\n\n- mixing total output with output above minimum;\n- applying startup\u002Fshutdown limits to the wrong quantity;\n- using the wrong `t` or `t-1` index;\n- over-constraining initial minimum up\u002Fdown obligations;\n- enforcing post-horizon obligations when the prompt excludes them;\n- treating cost curves as feasibility constraints;\n- choosing bad Big-M values;\n- requiring segment\u002Ftier variables when the trigger did not occur.\n\nDebug in stages: check shapes, relax one family at a time, add diagnostic slack variables, print largest violations, and compare local validation with final requirements.\n\n## Repair LPs And Heuristics\n\nA fixed-commitment repair LP is useful only if it includes every feasibility family judged in the final report. Do not repair only balance and capacity while omitting transition-dependent ramp, reserve, startup, shutdown, or minimum-duration limits.\n\nAlways rerun independent validation after repair.\n\n## Reporting Discipline\n\n- Use deterministic ordering and plain numeric values.\n- Do not include placeholder values.\n- Keep feasibility separate from proof quality\u002FMIP gap.\n- Use `null`\u002Fempty gap when no reliable bound exists, if the schema allows it.\n- Do not trust solver status or self-reported `\"pass\"` strings without validation.\n",{"name":5,"description":7},"main","cdcb6d230f3b15c3c33f8a2bdd16439f43602cbe","evaluation\u002Fresults\u002FWML\u002Ftasks\u002Fenergy-unit-commitment\u002Frepaired_skill\u002Fmilp-solver-workflow",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 milp-solver-workflow",{"tool":37,"command":38},"bunx","bunx skills add linyouxue\u002Fskill-repair-benchmark --skill milp-solver-workflow",{"tool":40,"command":41},"pnpm","pnpm dlx skills add linyouxue\u002Fskill-repair-benchmark --skill milp-solver-workflow","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fevaluation\u002Fresults\u002FWML\u002Ftasks\u002Fenergy-unit-commitment\u002Frepaired_skill\u002Fmilp-solver-workflow\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",[],[]]