| name | writing |
| description | Routing hub for writing that sounds human and stays coherent — Evan default voice pack, house anti-slop prose, technical articles, pure docs, and marketing handoff. Use when drafting or editing lessons, posts, threads, essays, docs, research notes, or anything that must not read like generic AI; when the user wants Evan's voice or a specific persona; or when the right writing skill is unclear. Routes to writer-style (default pack: evan), writing-prose, writing-technical, writing-docs, marketing. Prefer product-design for product UI craft; quality-check for ship/e2e.
|
| metadata | {"short-description":"Writing hub — Evan voice, prose, technical, docs"} |
Writing — routing hub
Sound like a person. Mean something. Stay correct.
Default long-form technical voice for this install: Evan Kim
(writer-style pack profiles/evan/), mined from
evan_writings.
LLM default prose fails three ways this hub blocks:
- Uniform polish — no human seams
- Fake opinion — wordy both-sides with no spine
- Style before facts — warm voice, wrong numbers
Route first → open one specialist → write. Max 1–3 skills.
writing
→ writer-style (evan default | other pack)
→ writing-prose (house human, no persona)
→ writing-technical (research / build-log form)
→ writing-docs (procedures / reference)
→ marketing/* (conversion frameworks only)
Routing table
| Your task | Skill | Load when |
|---|
| Sound like Evan (or “my voice” / default technical blog) | writer-style + pack evan | Lessons, research posts, build logs in his cadence |
| Named other author / new voice pack | writer-style + that pack | Persona work |
| House human tone, no persona | writing-prose | Docs-ish posts, edits, “less AI” |
| Research / empirical / mechanism explainer form | writing-technical | Structure + numbers; optional evan voice after |
| README, runbook, API reference, how-to | writing-docs | Followable procedures |
| Offer / StoryBrand / ads / viral | marketing | Conversion, not voice |
| Unclear multi-step writing | start here | Default |
Pipelines
A. Evan voice technical post
- Pick presentation mode (job, not topic) — see
writer-style/profiles/evan/modes.md
- writing-technical for that mode’s skeleton (or outline yourself)
- Fact-sheet (voice OFF)
- writer-style pack evan — load mode exemplars per
ROUTING.md; restyle, don’t re-derive
- Naturalness + deslop + fact diff (
validate_voice.py if available)
B. Pure docs
- writing-docs only
- Optional light writing-prose if prose got corporate
C. Marketing asset
- marketing frameworks first
- Optional writing-prose so it doesn’t sound like LLM brochureware
- Do not force Evan research-voice onto ads unless asked
Shared principles
- Facts first on technical claims
- Naturalness over polish (uneven rhythm, real seams)
- Opinion with a spine (costed pick)
- Paragraph dependence (no reshuffle-safe modules)
- Deslop subtractive (keep hard specifics)
- Original, not impersonation (primary idiolect only)
Evan default (quick)
Constant spine (all modes): structure-forward · payload early · evidence objects ·
insight as reframe · open gaps · no soft conclusion loops.
Mode by job (not topic): findings-note · field-log · claim-diary · systems-essay ·
bake-off · walkthrough · concept-map · design-brief · forum-fragment.
| Job | Mode |
|---|
| Measured result | findings-note |
| Production scars / build log | field-log |
| Paper / theory takeaways | claim-diary |
| Design A vs B argument | systems-essay |
| Same task, many tools | bake-off |
| Narrative how-to | walkthrough |
| Pure runbook | writing-docs |
| Lens / representation | concept-map |
| Artifact + constraints | design-brief |
| Unfinished public Q | forum-fragment |
Full spine: writer-style/profiles/evan/evan.md · modes: .../modes.md.
Sources (attribution)
See ATTRIBUTION.md.
When not to use this hub
| Task | Use instead |
|---|
| Product UI craft | product-design |
| Implement SPA | frontend-design |
| Prove / e2e | quality-check |
| Data pipelines | data |
Done criteria