sepia: De-AI writing at the layer that actually gives AI away (3.1k stars)
What it does
All text below is quoted directly from the project README.
>De-AI writing at the layer that actually gives AI away. Fiction gets its narrative architecture repaired before anyone touches word choice; professional documents (release notes, PR replies, postmortems, tickets, technical articles) each get rules matched to their venue.
>A portable Agent Skill: any agent that speaks the standard can load it, and the Skills CLI, which supports 77+ agents, installs it with one command. Claude Code, Codex, Grok Build, Antigravity, and QwenPaw additionally get native plugin packaging. One canonical SKILL.md, no per-platform forks. Four operations: write, review (diagnose only), refactor (minimal edits), recreate (full rewrite).
Why another humanizer
Every popular humanizer edits word choice and syntax. StoryScope (Russell et al., 2026: 61,608 stories, human + 5 frontier LLMs) showed that a classifier using narrative-structure features alone detects AI fiction at 93.2% macro-F1. In the same study's LAMP-edited condition, where human editors had rewritten the surface style, detection dropped only from 95.5% to 93.9%. The tells that survive are architectural: themes explained by the narrator, single-track causally-tidy plots, emotions rendered only as bodily sensation, no real-world references, no reader, linear time, endings resolved by protagonist growth and acceptance.
sepia turns those measured gaps, together with the related studies digested in research/, into a three-pass writing and revision protocol for fiction:
|Pass|Layer|Examples|
|:-|:-|:-|
|1|Narrative architecture (fiction)|stop explaining the theme, loosen the causal chain, back-load revelations, mix emotion modes, sparse character networks, name real things|
|2|Discourse flow|de-template the paragraph-question sequence, fix the mid-story sag, vary rhythm and positions|
|3|Surface style|the classic layer: clichés, syntax templates, vocabulary, register|
Each document type gets a thin rule file on top of one shared checklist:
|Domain|The gist|
|:-|:-|
|Release notes / announcements|user impact first, artifacts per claim, no marketing inflation|
|PR / issue replies|answer first, cite file:line, no reflex praise, length ∝ stakes|
|Postmortems|blameless toward people, merciless toward mechanisms; timestamps, dead ends, owned action items|
|Tickets / work orders|title = outcome, testable acceptance criteria, link don't repeat|
|Technical articles|open at the problem, one real dead end, one committed opinion, numbers with conditions|
|Long-form journalism (features, investigations, data stories)|lead and body in two registers, quotations keep their spoken texture, every number carries a comparison, no summary ending|
>Governing principle: Calibrate to the human distribution, don't invert the AI one. Humans sit at moderate values; a story with every rule applied is a new fingerprint. The skill selects 3–5 moves per story and leaves slack.
Operation entries
|Operation|Meaning|
|:-|:-|
|write|Create new prose|
|review|Diagnose without editing|
|refactor|Make minimal in-place edits|
|recreate|Rewrite from the source facts and intent|
|hemingway|Write or refactor fiction with the built-in Hemingway voice applied|
Sentence rhythm and Chinese calibration
>The style pass checks the spread of sentence lengths, the one syntactic measure on which every study that measured it agrees: human text varies more within a passage, in English and in Chinese.
Install
>Any agent (Skills CLI, 77+ agents): npx skills add Nanako0129/sepia -g
Repository
Nanako0129/sepia
Repository stats
|Metric|Value|
|:-|:-|
|Stars|3,147|
|Forks|196|
|Last release|v0.12.3 (2026-10-06)|
|Languages|Python 100%|
|Contributors|7|
|AI contributor accounts (Claude / Codex / Cursor / Copilot)|claude (210 commits)|
|AI co-authored commits|Claude: 211|
Main contributor
|Field|Value|
|:-|:-|
|Name|Nyanako|
|GitHub|@Nanako0129|
|Commits to this repo|223 (top contributor)|
|Bio|SRE · UX has frontend engineers. DX has SRE.|
|Company / Location|Taiwan|
|X / Twitter|@Nyanako0129|
|Followers|239|
|Public repos|58|
|On GitHub since|2018-11-04|
27 points
Mentions
以后你写的东西不再一眼 AI 腔，开口就是人味儿。 发现开源项目 sepia：去 AI 腔 Skill，据说三轮优化+30 项人工清单，专治讲太透、收尾太规整。 链接：Nanako0129/sepia 每天分享能帮你赚钱的 AI 工具，关注我不迷路 #AI工具 #AI写作 t.co
AI写的稿子一眼假？这个开源工具分3轮改。Sepia不换词，直接调叙事，还带30项人工检查清单，专治主题讲太透、结尾太规整。我试了一篇，改完确实像人写的。你们写东西最烦哪个AI味？ 仓库：Nanako0129/sepia t.co
Compatible with over 77 agents via the Skills CLI, sepia rewrites AI text by fixing underlying narrative structure instead of surface style. Nanako0129/sepia
sepia: De-AI writing at the layer that actually gives AI away (3.1k stars) What it does All text below is quoted directly from the project README. >De-AI writing at the layer that actually gives AI away. Fiction gets its narrative architecture repaired before anyone touches word choice; professional documents (release notes, PR replies, postmortems, tickets, technical articles) each get rules matched to their venue. >A portable Agent Skill: any agent that speaks the standard can load it, and the Skills CLI, which supports 77+ agents, installs it with one command. Claude Code, Codex, Grok Build, Antigravity, and QwenPaw additionally get native plugin packaging. One canonical SKILL.md, no per-platform forks. Four operations: write, review (diagnose only), refactor (minimal edits), recreate (full rewrite). Why another humanizer Every popular humanizer edits word choice and syntax. StoryScope (Russell et al., 2026: 61,608 stories, human + 5 frontier LLMs) showed that a classifier using narrative-structure features alone detects AI fiction at 93.2% macro-F1. In the same study's LAMP-edited condition, where human editors had rewritten the surface style, detection dropped only from 95.5% to 93.9%. The tells that survive are architectural: themes explained by the narrator, single-track causally-tidy plots, emotions rendered only as bodily sensation, no real-world references, no reader, linear time, endings resolved by protagonist growth and acceptance. sepia turns those measured gaps, together with the related studies digested in research/, into a three-pass writing and revision protocol for fiction: |Pass|Layer|Examples| |:-|:-|:-| |1|Narrative architecture (fiction)|stop explaining the theme, loosen the causal chain, back-load revelations, mix emotion modes, sparse character networks, name real things| |2|Discourse flow|de-template the paragraph-question sequence, fix the mid-story sag, vary rhythm and positions| |3|Surface style|the classic layer: clichés, syntax templates, vocabulary, register| Each document type gets a thin rule file on top of one shared checklist: |Domain|The gist| |:-|:-| |Release notes / announcements|user impact first, artifacts per claim, no marketing inflation| |PR / issue replies|answer first, cite file:line, no reflex praise, length ∝ stakes| |Postmortems|blameless toward people, merciless toward mechanisms; timestamps, dead ends, owned action items| |Tickets / work orders|title = outcome, testable acceptance criteria, link don't repeat| |Technical articles|open at the problem, one real dead end, one committed opinion, numbers with conditions| |Long-form journalism (features, investigations, data stories)|lead and body in two registers, quotations keep their spoken texture, every number carries a comparison, no summary ending| >Governing principle: Calibrate to the human distribution, don't invert the AI one. Humans sit at moderate values; a story with every rule applied is a new fingerprint. The skill selects 3–5 moves per story and leaves slack. Operation entries |Operation|Meaning| |:-|:-| |write|Create new prose| |review|Diagnose without editing| |refactor|Make minimal in-place edits| |recreate|Rewrite from the source facts and intent| |hemingway|Write or refactor fiction with the built-in Hemingway voice applied| Sentence rhythm and Chinese calibration >The style pass checks the spread of sentence lengths, the one syntactic measure on which every study that measured it agrees: human text varies more within a passage, in English and in Chinese. Install >Any agent (Skills CLI, 77+ agents): npx skills add Nanako0129/sepia -g Repository Nanako0129/sepia Repository stats |Metric|Value| |:-|:-| |Stars|3,147| |Forks|196| |Last release|v0.12.3 (2026-10-06)| |Languages|Python 100%| |Contributors|7| |AI contributor accounts (Claude / Codex / Cursor / Copilot)|claude (210 commits)| |AI co-authored commits|Claude: 211| Main contributor |Field|Value| |:-|:-| |Name|Nyanako| |GitHub|@Nanako0129| |Commits to this repo|223 (top contributor)| |Bio|SRE · UX has frontend engineers. DX has SRE.| |Company / Location|Taiwan| |X / Twitter|@Nyanako0129| |Followers|239| |Public repos|58| |On GitHub since|2018-11-04|
@vicaut @poteto @kr0der Use Nanako0129/syrtis I only use Grok and Grok Bot on this laptop, so all those models are through Cursor via Grok Bot. t.co
AI 写的东西为什么一眼假？ 6 万篇研究：人把 AI 小说字句全改一遍，检测率只从 95.5% 降到 93.9%——破绽在结构。 开源 skill sepia 照此去 AI 味：中文校准，实测句长波动只有人一半、零感叹号，一条命令装 77+ 款 agent。 3,007 星。 Nanako0129/sepia #AI写作 #Humanizer #开源 t.co