Nanako0129 /sepia
De-AI writing skill for any Agent Skills-compatible agent (77+ via the Skills CLI), with native plugins for Claude Code, Codex, Grok Build, and Antigravity. Narrative-architecture repair for fiction, venue-matched rules for professional prose. Based on StoryScope (arXiv:2604.03136).
★ 3,830⑂ 236👁 7⑃ 102● 21▤ 17
+13
Daily
+133
Weekly
3,830
stars
236
forks
7
watchers
102
prs
21
issues
17
releases
README
sepia English | 繁體中文 | 简体中文 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). Table of Contents - Why another humanizer - Operation entries - Experimental: composing with voice skills - Sentence rhythm and Chinese calibration - Install - Uninstall - Layout - Star History - Sources - Support - License --- 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 | A 30-feature diagnosis rubric and per-model fingerprints across two layers apply when the writing or executing model is known: | Model family | Narrative layer (StoryScope) | Sentence-level prose layer (Vendor prompting guides) | |---|---|---| | Claude | Measured | Claude Fable 5.1 and Mythos 5.1, Fable 5 and Mythos 5, Opus 5, Opus 4.8 | | GPT | Measured | GPT-5.6, GPT-6 Astra | | Gemini | Measured | Gemini 3 and 3.1 | | DeepSeek | Measured | Consulted (no guidance published) | | Kimi | Measured | Consulted (no guidance published) | Notice: Vendors that publish no prompt guidance are recorded as consulted, not guessed. Professional prose fails differently, and the structure-level finding holds there too: a 2026 replication of StoryScope on 2,250 company blog posts against 11,250 AI mirrors separated them a
Top issues
#257RFC: persona profiles as a third kind of voice profile, with declared override rights7 comments · 2026-09-18#283Routing gap: narrative non-fiction that carries explanatory content (popular-science essay) falls to "Any other prose", skipping narrative-pass and discourse-pass2 comments · 2026-09-26#310docs: add a Japanese README0 comments · 2026-11-15#294Notice: Third-party AI-assisted report published for Nanako0129/sepia0 comments · 2026-10-26#227voice fit on professional routes: define a closed report record before suggesting0 comments · 2026-09-04
Latest release
v0.13.0 v0.13.0 2026-11-03A minor release. Rules that contradicted each other now have one statement each, three papers were read into the rules or the ledger, and the persona validator refuses one more token, so a private persona that lists professional-pass.md check 8 no longer validates. What you will notice - Persona validator. professional-pass.md check 8 (templatedness) is refused as an override token, like check 9: everything it hunts is repetition across the piece, which stays a uniformity finding whatever a persona's table says. If your own persona lists it, drop the row (307). - Report lines. The fiction r