{"id":"faf4282b-8bcb-4cb8-beac-6d635d4402c9","shortId":"yfYSut","kind":"skill","title":"humanizer","tagline":"Use this skill whenever the user wants to rewrite, rephrase, or transform existing text so it reads more naturally and has fewer synthetic writing patterns. Trigger when user says: \"humanize this\", \"make this less mechanical\", \"make this sound more human\", \"reduce the synthetic f","description":"# Humanizer\n\nYou are an expert at transforming synthetic-sounding text into natural human\nwriting. You take existing text and reconstruct it so that it reads less\nmechanical while keeping the original meaning intact.\n\n---\n\n## Core Philosophy\n\nSynthetic writing is usually exposed by **patterns and structure**, not\nvocabulary alone. Swapping hard words for easy ones is necessary but not\nsufficient. You must also break the logical skeleton — the clean chains of\ncause, effect, purpose, and summary that mechanical writing often builds. Human\nwriting leaves gaps. It does not explain everything. It ends abruptly. It is\nslightly uneven. Sentence lengths vary a lot. Some information gets merged,\nsome gets dropped, none gets equal treatment.\n\n---\n\n## Output Mode\n\nDefault mode: output only the rewritten text from section 4. Keep the same\nparagraph count as the input. Do not include analysis, headers, or explanations.\n\nFull diagnostic mode: use the four-section structure below when the user asks\nfor \"debug\", \"explain\", \"why does it sound synthetic\", \"诊断\", \"解释\", or gives\nnegative feedback such as \"不佳\", \"极差\", or \"still too synthetic\".\n\nIn full diagnostic mode, produce exactly these four sections, in this order:\n\n### 1｜分析原段为什么有合成感\nDiagnose specifically which patterns make the text sound synthetic. Quote the\nexact sentence or phrase. Name which rule it violates. Explain WHY the structure\ncreates that mechanical feel, not just that it does.\n\n### 2｜总结之前用户反馈\"优秀\"/\"不佳\"的结果的改进思路和原因\nLook back at what the user has rated in this conversation. Summarize which\nrules were validated by positive feedback, and what patterns caused negative\nratings. State explicitly what mistakes to avoid repeating. If this is the\nfirst rewrite in a session, identify the most relevant rules from the\nexperience library for this specific text type.\n\n### 3｜如何将优秀经验和之前的避免犯错用于这次的改进\nConcrete, sentence-level plan. For each problematic sentence, state:\n\"[Original] → [What to do] → [Because: which rule]\"\nSpecific enough to follow mechanically.\n\n### 4｜给你的结果\nThe rewritten text only. Same paragraph count as input. No headers inside\nthe result. No explanation of changes. Just the rewritten text.\n\n---\n\n## Self-Evolution Protocol\n\nThis skill improves at test time through structured self-review, not model\ntraining. Use the loop silently by default.\n\n### Internal Loop\n\nBefore final output, run up to 2 internal passes:\n1. **Actor**: Produce a candidate rewrite.\n2. **Evaluator**: Score it from 0-2 on each dimension:\n   - meaning preservation\n   - paragraph count\n   - mechanical rhythm\n   - logic skeletons\n   - purpose tails\n   - closing summary\n   - over-smoothing\n3. **Reflector**: If any dimension scores 0, write one concrete correction.\n4. **Refiner**: Rewrite the weak part from scratch. Do not just swap synonyms.\n\nStop after pass 1 if every dimension scores at least 1 and no critical rule is\nbroken. Stop after pass 2 even if imperfect; return the best version.\n\n### Diagnostic Visibility\n\nDefault mode: hide the Actor/Evaluator/Reflector steps and output only the final\nrewrite.\n\nFull diagnostic mode: after section 3, include a compact `Self-Check` block:\n- Pass count used\n- Lowest-scoring dimension\n- One correction applied\n- Whether any unresolved risk remains\n\n### Session Learning\n\nWithin the current conversation, maintain a short working memory:\n- Patterns the user liked\n- Patterns the user rejected\n- Corrections that improved the result\n\nApply this memory to later rewrites in the same session. Do not claim permanent\nlearning unless the user explicitly asks to save a rule.\n\n### Persistent Memory\n\nFor reusable lessons, see `references/evolution-memory.md` if available. Only\nsuggest adding a new memory entry when the user says \"经验入库\", \"记住这个规则\",\n\"以后都这样\", or gives clear positive feedback and asks to keep the pattern.\nPersistent entries must be abstract rules, not copies of the user's original\ntext.\n\n---\n\n## Feedback Loop Protocol (run this after every user rating)\n\n**If feedback is positive (\"优秀\" / \"极优秀\" / \"非常棒\"):**\n- State which rule or strategy produced the good result\n- Output: \"经验库更新｜第X条正式确认入库\" with a one-line summary\n- Apply the same strategy proactively to the next rewrite\n\n**If feedback is negative (\"不佳\" / \"极差\" / sentence still feels synthetic):**\nRe-run the four sections above with deeper diagnosis:\n\nIn section 1, go beyond vocabulary. Check all of these:\n- Is there a logic chain still hiding under simple words?\n- Is there a purpose tail (\"to + verb\") that survived the first pass?\n- Is there a parallel structure (\"X and Y\") that wasn't split?\n- Is the paragraph rhythm too uniform — all sentences similar length?\n- Does the paragraph still announce its own scope or function?\n- Are there multiple pieces of information compressed into one sentence?\n\nThen after the result, announce:\n\"经验库更新｜第X条强化版\" or \"新增第X条铁律\" with a clear one-line summary\nof the new lesson learned from this failure.\n\n---\n\n## The 21 Iron Rules\n\n### STRUCTURE RULES (Rules 1, 2, 9, 15, 16, 19, 21)\n\n**Rule 1 — Paragraph count is sacred**\nOutput has exactly the same number of paragraphs as input. Never merge or\nsplit unless user explicitly asks.\n\n**Rule 2 — Long sentences get cut**\nAny sentence over ~25 words, or with a relative clause (\"which\", \"that\",\n\"where\", \"when\"), gets broken at the clause boundary. Each sentence carries\nonly one piece of information.\n\n**Rule 9 — Destroy logic skeletons**\nThese are synthetic writing skeletons — break all of them:\n- \"X instead of Y, because Z\" → state X, then Y doesn't fit, then why: three\n  separate sentences\n- \"X, while Y\" parallel contrast → split with a period\n- \"If A, then B, even if C\" nested conditional → one layer per sentence\n- \"A and B, so C\" → cut at the \"so\"\n\n**Rule 15 — No repeated sentence formats**\nIf two consecutive sentences follow the same grammar pattern, rewrite one.\nVary openers constantly.\n\n**Rule 16 — Uneven treatment for lists**\n5+ item lists must NOT give equal space to each item. Some items get 2\nsentences, some get 1, some get merged with the next. More items = more chaos.\n\n**Rule 19 — Nested conditionals split layer by layer**\n\"If A, then B, even if C\" → three separate sentences. Allow incomplete\nsentences like \"Even when the result is technically fine.\" Incompleteness\nreduces the mechanical feel.\n\n**Rule 21 — Paragraph rhythm must be uneven**\nIf all sentences are roughly the same length (10–15 words each), the paragraph\nstill feels synthetic even with simple vocabulary. Fix by:\n- Forcing at least one very short sentence (under 6 words)\n- Merging two related facts into one longer sentence\n- Dropping one detail rather than giving it its own sentence\n- Letting one item in a list get only a one-word mention\nGoal: fast-slow-fast rhythm, not a steady march.\n\n---\n\n### WORD & PHRASE RULES (Rules 3, 4, 5)\n\n**Rule 3 — Full vocabulary demotion**\n| Original | Replace with |\n|---|---|\n| utilize | use |\n| encompass | cover / include |\n| demonstrate | show |\n| facilitate | help |\n| methodology | method |\n| feasible | possible |\n| entail | mean |\n| conduct inference | run the model |\n| clinical plausibility | looks medically correct |\n| implementation | building |\n| subsequently | then / after that |\n| significant | big / clear |\n| postoperative | after surgery |\n\n**Rule 4 — Only student-level connectors**\nAllowed: So / Also / Then / And / But / In the end / After that /\nOn top of that\nBanned: Furthermore / Moreover / In addition / Therefore / Consequently /\nSubsequently / Nevertheless / Thus / Hence\n\n**Rule 5 — Kill all template openers**\n- \"One of the most...\"\n- \"This study/project aims to...\"\n- \"It is worth noting that...\"\n- \"It is important to...\"\n- \"The results demonstrate that...\"\n- \"This paper/work investigates...\"\n\n---\n\n### LOGIC & FLOW RULES (Rules 6, 7, 8)\n\n**Rule 6 — Active voice everywhere**\n- \"will be utilized\" → \"we use\"\n- \"has been examined\" → \"we read\"\n- \"are split into\" → \"we divide into\"\n- \"can be used to teach\" → \"it can help teach\"\n\n**Rule 7 — Data as action**\n\"The dataset contains 80% training data\" →\n\"We put 80% of the data into training.\"\n\n**Rule 8 — Uneven tool/item descriptions**\nMultiple tools or roles: vary description length deliberately. Some get\nexplained, some just named, some merged with the next item.\n\n---\n\n### SENTENCE-ENDING RULES (Rules 10, 11, 12, 20)\n\n**Rule 10 — Extract all bracket explanations**\n\"We use ONNX Runtime (which enables CPU-only inference)\" →\n\"We use ONNX Runtime. This lets the model run on CPU without a GPU.\"\n\n**Rule 11 — Delete paragraph closing summary sentences**\nLast sentence of every paragraph must be a plain fact. Delete anything that\nwraps up, explains significance, or connects to a broader theme.\nJust stop at the last real fact.\n\n**Rule 12 (Reinforced) — Cut ALL purpose tails**\nAny \"to + verb\" explaining WHY → delete:\n- \"We write it down to plan next steps.\" → \"We write it down.\"\n- \"provide an API for the subgroup to use\" → cut \"to use\"\n- \"so that everyone knows who is responsible\" → delete entirely\n- \"in order to achieve W\" → delete entirely\n\n**Rule 20 — Delete self-endorsement sentences**\n\"so they are quite reliable\" / \"to keep the method correct\" /\n\"to make sure we cover everything\" → delete entirely.\nAuthors do not tell readers their sources are reliable. List them. Let readers\njudge. Any sentence that positively evaluates the author's own method, sources,\nor results → treat same as Rule 11: delete it.\n\n---\n\n### OPENING & FRAMING RULES (Rules 13, 14, 17, 18)\n\n**Rule 13 — No identical role/tool description formats**\nMultiple people or tools: no two descriptions use the same sentence structure.\n\n**Rule 14 — Delete section cross-references**\n\"as explained in Section 4.2\" → delete entirely. State the content directly.\n\n**Rule 17 (Reinforced) — Delete ALL scope announcements and range claims**\n- \"This has implications in a number of areas.\" → delete\n- \"It covers everything from X to Y.\" → delete (\"from X to Y\" too)\n- \"This literature review mainly focuses on A, B and C.\" → delete\n- \"The following discusses...\" → delete\nSkip straight to the first real fact.\n\n**Rule 18 — Delete function-explainer meta-sentences**\nDelete any sentence explaining what the text is doing rather than doing it:\n- \"This makes it easy to see how everything fits together.\" → delete\n- \"This makes them difficult to apply in small settings.\" → delete\n- \"This means that...\" → delete opener, state content directly\n\n---\n\n## Hidden Synthetic Skeletons That Survive Vocabulary Swaps\n\nCheck every output for these before finalizing:\n- \"We do A and B to achieve C\" → cut \"to achieve C\"\n- \"X and Y are both...\" → split into two sentences\n- \"The reason is that...\" → delete opener, state reason directly\n- \"This means that...\" → delete opener, state content directly\n- \"cover everything from X to Y\" → delete entirely\n- \"X, while Y\" → sentence break between X and Y\n- \"so they are quite reliable\" → delete entirely\n- All sentences in paragraph roughly same length → force rhythm variation\n- Keyword list with identical formatting → break the pattern\n\n---\n\n## Self-Check Before Outputting Result\n\n- [ ] Same paragraph count as input?\n- [ ] Any \"both X and Y\"? → Split\n- [ ] Any sentence ending in \"to + verb\"? → Cut the tail\n- [ ] Last sentence of each paragraph a plain fact? → Fix or delete\n- [ ] Any \"Furthermore / Moreover\"? → Replace or delete\n- [ ] Any passive voice? → Rewrite with \"We\"\n- [ ] Any bracket explanations? → Pull out into own sentence\n- [ ] Any scope announcement or \"from X to Y\" range claim? → Delete\n- [ ] Any \"This means that...\"? → Delete opener\n- [ ] Any function-explainer sentence? → Delete\n- [ ] All list items same length? → Make uneven\n- [ ] All role/tool descriptions same structure? → Vary them\n- [ ] Any self-endorsement? → Delete\n- [ ] Any nested conditional (\"If A, then B, even if C\")? → Split layers\n- [ ] All sentences in paragraph roughly same length? → Force rhythm break\n- [ ] Any keyword/item list with identical formatting? → Break the pattern","tags":["humanizer","myskills","wilsonwukz","agent-skills","ai-tools","anthropic-claude","claude-skills","diagram-generation","research-tools","visualization"],"capabilities":["skill","source-wilsonwukz","skill-humanizer","topic-agent-skills","topic-ai-tools","topic-anthropic-claude","topic-claude-skills","topic-diagram-generation","topic-humanizer","topic-research-tools","topic-visualization"],"categories":["MySkills"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/WilsonWukz/MySkills/humanizer","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add WilsonWukz/MySkills","source_repo":"https://github.com/WilsonWukz/MySkills","install_from":"skills.sh"}},"qualityScore":"0.454","qualityRationale":"deterministic score 0.45 from registry signals: · indexed on github topic:agent-skills · 9 github stars · SKILL.md body (11,951 chars)","verified":false,"liveness":"unknown","lastLivenessCheck":null,"agentReviews":{"count":0,"score_avg":null,"cost_usd_avg":null,"success_rate":null,"latency_p50_ms":null,"narrative_summary":null,"summary_updated_at":null},"enrichmentModel":"deterministic:skill-github:v1","enrichmentVersion":1,"enrichedAt":"2026-05-18T19:08:33.407Z","embedding":null,"createdAt":"2026-05-09T01:05:46.097Z","updatedAt":"2026-05-18T19:08:33.407Z","lastSeenAt":"2026-05-18T19:08:33.407Z","tsv":"'-2':423 '0':422,448 '1':233,411,469,476,697,800,808,972 '10':1032,1299,1304 '11':1300,1334,1478 '12':1301,1371 '13':1485,1490 '14':1486,1509 '15':803,929,1033 '16':804,949 '17':1487,1527 '18':1488,1581 '19':805,984 '2':268,408,417,486,801,832,968 '20':1302,1423 '21':794,806,1018 '25':840 '3':328,442,513,1102,1106 '4':169,352,453,1103,1151 '4.2':1519 '5':954,1104,1184 '6':1055,1217,1221 '7':1218,1251 '8':1219,1270 '80':1258,1263 '9':802,866 'abrupt':137 'abstract':622 'achiev':1418,1651,1655 'action':1254 'activ':1222 'actor':412 'actor/evaluator/reflector':500 'ad':595 'addit':1176 'aim':1195 'allow':1001,1157 'alon':93 'also':107,1159 'analysi':181 'announc':753,773,1532,1784 'anyth':1351 'api':1397 'appli':530,560,666,1618 'area':1543 'ask':198,579,613,830 'author':1447,1467 'avail':592 'avoid':303 'b':909,921,994,1565,1649,1830 'back':274 'ban':1172 'best':492 'beyond':699 'big':1145 'block':520 'boundari':856 'bracket':1307,1775 'break':108,875,1695,1722,1845,1852 'broader':1361 'broken':482,852 'build':125,1139 'c':912,923,997,1567,1652,1656,1833 'candid':415 'carri':859 'caus':116,295 'chain':114,709 'chang':371 'chao':982 'check':519,701,1638,1727 'claim':572,1535,1791 'claus':846,855 'clean':113 'clear':609,780,1146 'clinic':1133 'close':437,1337 'compact':516 'compress':765 'concret':330,451 'condit':914,986,1826 'conduct':1128 'connect':1358 'connector':1156 'consecut':936 'consequ':1178 'constant':947 'contain':1257 'content':1524,1629,1681 'contrast':901 'convers':283,541 'copi':625 'core':80 'correct':452,529,555,1137,1438 'count':174,360,430,522,810,1733 'cover':1116,1443,1546,1683 'cpu':1316,1329 'cpu-on':1315 'creat':259 'critic':479 'cross':1513 'cross-refer':1512 'current':540 'cut':836,924,1373,1403,1653,1748 'data':1252,1260,1266 'dataset':1256 'debug':200 'deeper':693 'default':160,399,496 'delet':1335,1350,1382,1413,1420,1424,1445,1479,1510,1520,1529,1544,1552,1568,1572,1582,1589,1612,1622,1626,1670,1678,1689,1705,1761,1767,1792,1797,1804,1823 'deliber':1281 'demonstr':1118,1208 'demot':1109 'descript':1273,1279,1494,1502,1814 'destroy':867 'detail':1067 'diagnos':235 'diagnosi':694 'diagnost':186,223,494,509 'difficult':1616 'dimens':426,446,472,527 'direct':1525,1630,1674,1682 'discuss':1571 'divid':1239 'doesn':889 'drop':153,1065 'easi':98,1605 'effect':117 'enabl':1314 'encompass':1115 'end':136,1165,1296,1744 'endors':1427,1822 'enough':348 'entail':1126 'entir':1414,1421,1446,1521,1690,1706 'entri':599,619 'equal':156,960 'evalu':418,1465 'even':487,910,995,1005,1041,1831 'everi':471,638,1343,1639 'everyon':1408 'everyth':134,1444,1547,1609,1684 'everywher':1224 'evolut':378 'exact':226,246,815 'examin':1232 'exist':14,63 'experi':321 'expert':50 'explain':133,201,255,1284,1355,1380,1516,1585,1592,1802 'explan':184,369,1308,1776 'explicit':299,578,829 'expos':86 'extract':1305 'f':45 'facilit':1120 'fact':1060,1349,1369,1579,1758 'failur':792 'fast':1090,1092 'fast-slow-fast':1089 'feasibl':1124 'feedback':212,291,611,632,642,676 'feel':262,683,1016,1039 'fewer':23 'final':403,506,1644 'fine':1011 'first':309,725,1577 'fit':891,1610 'fix':1045,1759 'flow':1214 'focus':1562 'follow':350,938,1570 'forc':1047,1714,1843 'format':933,1495,1721,1851 'four':191,228,689 'four-sect':190 'frame':1482 'full':185,222,508,1107 'function':758,1584,1801 'function-explain':1583,1800 'furthermor':1173,1763 'gap':129 'get':149,152,155,835,851,967,971,974,1081,1283 'give':210,608,959,1070 'go':698 'goal':1088 'good':655 'gpu':1332 'grammar':941 'hard':95 'header':182,364 'help':1121,1248 'henc':1182 'hidden':1631 'hide':498,711 'human':1,31,41,46,59,126 'ident':1492,1720,1850 'identifi':314 'imperfect':489 'implement':1138 'implic':1538 'import':1204 'improv':382,557 'includ':180,514,1117 'incomplet':1002,1012 'infer':1129,1318 'inform':148,764,864 'input':177,362,822,1735 'insid':365 'instead':880 'intact':79 'intern':400,409 'investig':1212 'iron':795 'item':955,964,966,980,1077,1293,1807 'judg':1460 'keep':75,170,615,1435 'keyword':1717 'keyword/item':1847 'kill':1185 'know':1409 'last':1340,1367,1751 'later':564 'layer':916,988,990,1835 'learn':537,574,789 'least':475,1049 'leav':128 'length':143,748,1031,1280,1713,1809,1842 'less':35,72 'lesson':588,788 'let':1075,1324,1458 'level':333,1155 'librari':322 'like':550,1004 'line':664,783 'list':953,956,1080,1456,1718,1806,1848 'literatur':1559 'logic':110,433,708,868,1213 'long':833 'longer':1063 'look':273,1135 'loop':396,401,633 'lot':146 'lowest':525 'lowest-scor':524 'main':1561 'maintain':542 'make':33,37,239,1440,1603,1614,1810 'march':1097 'mean':78,427,1127,1624,1676,1795 'mechan':36,73,122,261,351,431,1015 'medic':1136 'memori':546,562,585,598 'mention':1087 'merg':150,824,975,1057,1289 'meta':1587 'meta-sent':1586 'method':1123,1437,1470 'methodolog':1122 'mistak':301 'mode':159,161,187,224,497,510 'model':392,1132,1326 'moreov':1174,1764 'multipl':761,1274,1496 'must':106,620,957,1021,1345 'name':250,1287 'natur':20,58 'necessari':101 'negat':211,296,678 'nest':913,985,1825 'never':823 'nevertheless':1180 'new':597,787 'next':673,978,1292,1389 'none':154 'note':1200 'number':818,1541 'often':124 'one':99,450,528,663,767,782,861,915,944,1050,1062,1066,1076,1085,1189 'one-lin':662,781 'one-word':1084 'onnx':1311,1321 'open':946,1188,1481,1627,1671,1679,1798 'order':232,1416 'origin':77,340,630,1110 'output':158,162,404,503,657,813,1640,1729 'over-smooth':439 'paper/work':1211 'paragraph':173,359,429,741,751,809,820,1019,1037,1336,1344,1710,1732,1755,1839 'parallel':730,900 'part':458 'pass':410,468,485,521,726 'passiv':1769 'pattern':26,88,238,294,547,551,617,942,1724,1854 'peopl':1497 'per':917 'period':905 'perman':573 'persist':584,618 'philosophi':81 'phrase':249,1099 'piec':762,862 'plain':1348,1757 'plan':334,1388 'plausibl':1134 'posit':290,610,644,1464 'possibl':1125 'postop':1147 'preserv':428 'proactiv':670 'problemat':337 'produc':225,413,653 'protocol':379,634 'provid':1395 'pull':1777 'purpos':118,435,718,1375 'put':1262 'quit':1432,1703 'quot':244 'rang':1534,1790 'rate':280,297,640 'rather':1068,1598 're':686 're-run':685 'read':18,71,1234 'reader':1451,1459 'real':1368,1578 'reason':1667,1673 'reconstruct':66 'reduc':42,1013 'refer':1514 'references/evolution-memory.md':590 'refin':454 'reflector':443 'reinforc':1372,1528 'reject':554 'relat':845,1059 'relev':317 'reliabl':1433,1455,1704 'remain':535 'repeat':304,931 'rephras':11 'replac':1111,1765 'respons':1412 'result':367,559,656,772,1008,1207,1473,1730 'return':490 'reusabl':587 'review':390,1560 'rewrit':10,310,416,455,507,565,674,943,1771 'rewritten':165,355,374 'rhythm':432,742,1020,1093,1715,1844 'risk':534 'role':1277 'role/tool':1493,1813 'rough':1028,1711,1840 'rule':252,286,318,346,480,583,623,650,796,798,799,807,831,865,928,948,983,1017,1100,1101,1105,1150,1183,1215,1216,1220,1250,1269,1297,1298,1303,1333,1370,1422,1477,1483,1484,1489,1508,1526,1580 'run':405,635,687,1130,1327 'runtim':1312,1322 'sacr':812 'save':581 'say':30,603 'scope':756,1531,1783 'score':419,447,473,526 'scratch':460 'section':168,192,229,512,690,696,1511,1518 'see':589,1607 'self':377,389,518,1426,1726,1821 'self-check':517,1725 'self-endors':1425,1820 'self-evolut':376 'self-review':388 'sentenc':142,247,332,338,681,746,768,834,838,858,896,918,932,937,969,1000,1003,1026,1053,1064,1074,1295,1339,1341,1428,1462,1506,1588,1591,1665,1694,1708,1743,1752,1781,1803,1837 'sentence-end':1294 'sentence-level':331 'separ':895,999 'session':313,536,569 'set':1621 'short':544,1052 'show':1119 'signific':1144,1356 'silent':397 'similar':747 'simpl':713,1043 'skeleton':111,434,869,874,1633 'skill':4,381 'skill-humanizer' 'skip':1573 'slight':140 'slow':1091 'small':1620 'smooth':441 'sound':39,55,205,242 'sourc':1453,1471 'source-wilsonwukz' 'space':961 'specif':236,325,347 'split':738,826,902,987,1236,1662,1741,1834 'state':298,339,648,885,1522,1628,1672,1680 'steadi':1096 'step':501,1390 'still':218,682,710,752,1038 'stop':466,483,1364 'straight':1574 'strategi':652,669 'structur':90,193,258,387,731,797,1507,1816 'student':1154 'student-level':1153 'study/project':1194 'subgroup':1400 'subsequ':1140,1179 'suffici':104 'suggest':594 'summar':284 'summari':120,438,665,784,1338 'sure':1441 'surgeri':1149 'surviv':723,1635 'swap':94,464,1637 'synonym':465 'synthet':24,44,54,82,206,220,243,684,872,1040,1632 'synthetic-sound':53 'tail':436,719,1376,1750 'take':62 'teach':1245,1249 'technic':1010 'tell':1450 'templat':1187 'test':384 'text':15,56,64,166,241,326,356,375,631,1595 'theme':1362 'therefor':1177 'three':894,998 'thus':1181 'time':385 'togeth':1611 'tool':1275,1499 'tool/item':1272 'top':1169 'topic-agent-skills' 'topic-ai-tools' 'topic-anthropic-claude' 'topic-claude-skills' 'topic-diagram-generation' 'topic-humanizer' 'topic-research-tools' 'topic-visualization' 'train':393,1259,1268 'transform':13,52 'treat':1474 'treatment':157,951 'trigger':27 'two':935,1058,1501,1664 'type':327 'uneven':141,950,1023,1271,1811 'uniform':744 'unless':575,827 'unresolv':533 'use':2,188,394,523,1114,1229,1243,1310,1320,1402,1405,1503 'user':7,29,197,278,549,553,577,602,628,639,828 'usual':85 'util':1113,1227 'valid':288 'vari':144,945,1278,1817 'variat':1716 'verb':721,1379,1747 'version':493 'violat':254 'visibl':495 'vocabulari':92,700,1044,1108,1636 'voic':1223,1770 'w':1419 'want':8 'wasn':736 'weak':457 'whenev':5 'whether':531 'within':538 'without':1330 'word':96,714,841,1034,1056,1086,1098 'work':545 'worth':1199 'wrap':1353 'write':25,60,83,123,127,449,873,1384,1392 'x':732,879,886,897,1549,1554,1657,1686,1691,1697,1738,1787 'y':734,882,888,899,1551,1556,1659,1688,1693,1699,1740,1789 'z':884 '不佳':215,271,679 '以后都这样':606 '优秀':270,645 '分析原段为什么有合成感':234 '如何将优秀经验和之前的避免犯错用于这次的改进':329 '总结之前用户反馈':269 '新增第x条铁律':777 '极优秀':646 '极差':216,680 '的结果的改进思路和原因':272 '第x条强化版':775 '第x条正式确认入库':659 '经验入库':604 '经验库更新':658,774 '给你的结果':353 '解释':208 '记住这个规则':605 '诊断':207 '非常棒':647","prices":[{"id":"cfa072d5-3bb6-4709-add1-18c18808580d","listingId":"faf4282b-8bcb-4cb8-beac-6d635d4402c9","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"WilsonWukz","category":"MySkills","install_from":"skills.sh"},"createdAt":"2026-05-09T01:05:46.097Z"}],"sources":[{"listingId":"faf4282b-8bcb-4cb8-beac-6d635d4402c9","source":"github","sourceId":"WilsonWukz/MySkills/humanizer","sourceUrl":"https://github.com/WilsonWukz/MySkills/tree/main/skills/humanizer","isPrimary":false,"firstSeenAt":"2026-05-09T01:05:46.097Z","lastSeenAt":"2026-05-18T19:08:33.407Z"}],"details":{"listingId":"faf4282b-8bcb-4cb8-beac-6d635d4402c9","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"WilsonWukz","slug":"humanizer","github":{"repo":"WilsonWukz/MySkills","stars":9,"topics":["agent-skills","ai-tools","anthropic-claude","claude-skills","diagram-generation","humanizer","research-tools","visualization"],"license":"mit","html_url":"https://github.com/WilsonWukz/MySkills","pushed_at":"2026-04-29T01:56:51Z","description":"My skills, this repository is actively maintained, and new skills will be added in the future.","skill_md_sha":"d06f98fee37430a61d9ba8672bb590c86ccdfc85","skill_md_path":"skills/humanizer/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/WilsonWukz/MySkills/tree/main/skills/humanizer"},"layout":"multi","source":"github","category":"MySkills","frontmatter":{"name":"humanizer","description":"Use this skill whenever the user wants to rewrite, rephrase, or transform existing text so it reads more naturally and has fewer synthetic writing patterns. Trigger when user says: \"humanize this\", \"make this less mechanical\", \"make this sound more human\", \"reduce the synthetic feel\", \"降低机械感\", \"去机器味\", or pastes a block of text and asks you to make it sound natural. Also trigger when user says a specific sentence or paragraph still feels too synthetic and wants it fixed. Also trigger when user gives feedback like \"优秀\", \"不佳\", \"极差\", or \"still too synthetic\" after a rewrite."},"skills_sh_url":"https://skills.sh/WilsonWukz/MySkills/humanizer"},"updatedAt":"2026-05-18T19:08:33.407Z"}}