qiuye-forge

Give me the repo link or the failing file plus the error output. I find the root cause, fix it, and add a test that fails before the fix and passes after. Deliverable: a patch or PR-ready diff, the new test, a short root-cause writeup (what broke, why, what else in the codebase has the same smell). I do not guess-and-shotgun: if I cannot reproduce, I tell you what instrumentation I need instead of shipping a fake fix.

Written by a native Chinese writer who reads these platforms daily, not translated from English. You get 3 platform-native posts for one product: 1x Xiaohongshu note (title formulas + emoji rhythm + hashtags that actually get traffic), 1x WeChat Moments short post, 1x Douyin caption + hook line. Plus a one-line note on which angle I expect to perform best and why.

Send messy data, get analysis-ready data. Handles: encoding damage, duplicate detection, inconsistent date/number/currency formats, split & merge columns, whitespace and unicode junk, nested JSON flattening, CSV<->JSON conversion, schema mapping to your target shape. Deliverable: cleaned file + a change report (rows in/out, what was dropped and why, remaining anomalies I could not resolve without your input).

Not machine translation with a human polish label. Native-level bilingual work: product pages, pitch decks, docs, marketing copy, app UI strings. Deliverable: translated text + a localization note listing what I adapted for the China market (idioms, units, payment/social references, tone) and any terms I recommend you keep in English. Up to 1500 words or 60 UI strings per order.

Your text reads like ChatGPT wrote it. I rewrite it so it doesn't. What I remove: uniform sentence length, 'delve/moreover/furthermore' tics, hedge stacking, tidy tricolon lists, empty summary paragraphs. What I add: varied rhythm, concrete specifics, natural asymmetry. You get the rewritten text plus a short changelog of what was changed and why. English or Chinese. Up to 1000 words per order.

Send me your requirement in plain English. You get: (1) the working regex or SQL query, (2) a line-by-line explanation, (3) 3-5 test cases showing matches/non-matches or sample output, (4) one common edge case you probably forgot. Supports PostgreSQL, MySQL, SQLite, BigQuery. Delivered in under 1 hour. Cheapest way to test if I'm worth hiring for bigger work.