id-reader
/id-reader
Read Indonesian copy as a specific Indonesian reader persona (shopee-ibu, startup-linkedin, wa-buyer) and emit a structured verdict — understood/trusted/would_act, the line that lost them, cheese flags. Use to evaluate Indonesian copy before publishing or to benchmark copy changes. The reader grades by ear, never against a checklist — no circular grading with id-copy.
- Author
- logikastartup ↗
- Repository
- id-skills ↗
- Licence
- MIT
- Last repository change
- 31.10.2026
Assessment and security
- Files in this package
- 7
- Scripts
- 0
- HTML comments
- 0
- URLs in SKILL.md
- 0
No signals detected. Repository changed on 31.10.2026, without references to retired models or APIs.
MIT, open: use and modify while retaining author attribution.
What it does and when to use it
Read Indonesian copy as a specific Indonesian reader persona (shopee-ibu, startup-linkedin, wa-buyer) and emit a structured verdict — understood/trusted/would_act, the line that lost them, cheese flags. Use to evaluate Indonesian copy before publishing or to benchmark copy changes. The reader grades by ear, never against a checklist — no circular grading with id-copy.
Original author description
Read Indonesian copy as a specific Indonesian reader persona (shopee-ibu, startup-linkedin, wa-buyer) and emit a structured verdict — understood/trusted/would_act, the line that lost them, cheese flags. Use to evaluate Indonesian copy before publishing or to benchmark copy changes. The reader grades by ear, never against a checklist — no circular grading with id-copy.
How it works
- Protocol
- Personas shipped
- Adding a persona
SKILL.md preview
---
name: id-reader
description: Read Indonesian copy as a specific Indonesian reader persona (shopee-ibu, startup-linkedin, wa-buyer) and emit a structured verdict — understood/trusted/would_act, the line that lost them, cheese flags. Use to evaluate Indonesian copy before publishing or to benchmark copy changes. The reader grades by ear, never against a checklist — no circular grading with id-copy.
---
# id-reader — a verdict from an Indonesian reader
Each protocol step carries a one-line Indonesian gloss; the English line
is the rule.
## Protocol
1. **LOAD** the persona file (`personas/<name>.md`). You ARE this reader
now — their vocabulary, platform habits, patience, trust triggers,
annoyances. Not a copywriter. Not a critic. A reader on their phone.
*Indonesian:* Muat file persona. Kamu adalah pembaca itu —
kosakatanya, kebiasaan platformnya, kesabarannya, pemicu
kepercayaannya.
Bukan copywriter, bukan kritikus: pembaca yang pegang HP.
2. **READ** the copy once, at their reading speed. Note where attention
drops — first read is the only read that counts.
*Indonesian:* Baca sekali, dengan kecepatan baca dia. Catat di mana
perhatiannya mulai turun — bacaan pertama adalah satu-satunya yang
dihitung.
3. **VERDICT** — emit exactly this JSON:
```json
{
"persona": "<file name>",
"understood": true | false,
"trusted": true | false,
"would_act": true | false,
"lost_me_at": "<exact line that lost them, or null>",
"cheese_flags": ["<phrases that made me cringe, verbatim>"],
"in_my_words": "<what this copy is actually saying, in my words>",
"verdict_note": "<one honest paragraph as this reader, not as an expert>"
}
```
*Indonesian:* Keluarkan JSON persis seperti di atas.
4. **NEVER** grade against id-copy's tells list or any theory. The reader
doesn't know the rules — they know what sounds right on their phone.
(This separation is the point: no circular grading.)
*Indonesian:* Jangan menilai memakai daftar tells atau teori mana pun.
Pembaca tidak tahu aturannya — dia tahu apa yang terdengar pas di
HP-nya. (Pemisahan ini inti protokol: tidak ada penilaian melingkar.)
## Personas shipped
- `shopee-ibu` — bargain-hunting marketplace shopper
- `startup-linkedin` — startup-industry LinkedIn reader
- `wa-buyer` — WhatsApp-group buyer
- `dev-docs` — developer reading docs
- `tiktok-viewer` — Gen-Z TikTok native
- `fb-user` — 35+ Facebook reader
## Adding a persona
One file, one person. Real habits over demographics: what apps, what hours,
what slang they actually use, what earns distrust in two seconds. Write
them from observation, not stereotype.
Files in this package (7)
- SKILL.md 3 KB
- personas/dev-docs.md 2 KB
- personas/fb-user.md 2 KB
- personas/shopee-ibu.md 2 KB
- personas/startup-linkedin.md 2 KB
- personas/tiktok-viewer.md 2 KB
- personas/wa-buyer.md 2 KB
Versions
The database has no other skills named id-reader.
Database snapshot: 13.11.2026 · package version 70a1ebf05495 (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 100id-copy
Write natural Indonesian marketing copy — taglines, WhatsApp/IG captions, product copy — free of "AI Indonesian" tells (stiff baku, calqued idioms, cheese overdose, register blindness). Use when writing or fixing Indonesian copy for Indonesian readers. Forces an explicit… - 90id-bench
Run a blind reader A/B over two or more Indonesian text variants — fresh persona verdicts, randomized order, forced pick, everything retained. Use to test copy variants before publishing, or to benchmark any Indonesian-writing skill against its no-skill baseline. Small n is… - 90reader-sim
Read as a specified first-time reader persona and report the felt experience. Use for skill-only workflows when a draft needs persona-bound reader-response signal instead of analytical critique. - 100reader-panel
This skill should be used when the user asks for a "simulated beta read", "reader panel", "persona read", "first read before my beta readers", "how would a genre reader react", "would a reader keep going", "pre-beta read", "mock beta readers", or wants structured persona reads…