How do I install the Inbox skill?

Run npx skillmds@latest add dkadts/inbox in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Inbox safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Inbox?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Inbox free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

dkadts (@dkadts) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Inbox by dkadts — Agent Skill

Name: Inbox
Author: dkadts

File contents

/inbox

First pass over client materials. Does not convert (that's /media), does not analyze text (that's /copy-analyze) — only sorts and validates.

When to use

Client says "materials uploaded".
After a new batch arrives (top-up).

Process

Step 1. Read brief/08-content-pipeline.md Load requirements: naming, formats, sizes.

Step 2. Walk inbox/ folders

For each folder (texts/, media/photos/, media/videos/, brand/logo/, brand/fonts/, notes/):

List files.
Verify: naming ASCII? correct format? size within limits?
Mark: ✅ OK / ⚠️ warning / ❌ reject.

Step 3. Specific checks

texts/:

UTF-8?
Section headings present (# Hero, # Offer)?

media/photos/:

Min side ≥ 2400px (read via image tool or identify).
sRGB (not Adobe RGB — otherwise color shift in the browser).
ASCII filenames.
No IMG_1234.jpg.

media/videos/:

1080p+.
Bitrate ≤ 20 Mbps.
H.264 in MP4 (not HEVC — poor browser support).

brand/logo/:

SVG only.
Text outlined (check: is there a <text> tag inside SVG — if yes, NOT outlined).
No embedded raster (no <image> inside).

brand/fonts/:

.woff2 only.
LICENSE.txt present.

notes/:

List with type (audio / image / md / other).
Transcribe voice notes (if possible).

Step 4. Produce the report

Create/update inbox/inbox-report.md:

# Inbox report — 2026-08-08

## Overall
- Files total: N
- OK: N | Warning: N | Reject: N

## texts/
- `copy.md` — ✅ OK (UTF-8, 5 sections)
- `about.docx` — ⚠️ convert to MD

## media/photos/
- `hero-01.jpg` — ✅ 3600×2400, sRGB
- `IMG_1234.jpg` — ❌ rename to ASCII
- `founder.tif` — ⚠️ 6000×4000 > 6000px — will be cropped

## brand/logo/
- `logo.svg` — ❌ text not outlined (found <text>)

## Summary for the client
### Ready
- Texts can go to /copy-analyze
- Hero photo (after rename)

### Needs rework
- 3 photos: rename
- Logo: outline (in Illustrator: Type → Create Outlines)
- `about.docx`: convert to MD

## Next
- Wait for client rework
- OR if user says "ship with what we have" — run /media on ok-files and /copy-analyze on texts

Step 5. Extract links from notes/

If notes contain URLs — offer to move them into references/ by aspect.

Rules

Do not convert files here. Only validate and sort.
Do not complain if materials are thin. Just report "we have X, for a full landing we need Y".
Do not delete reject files. Only mark in the report — user decides.

/copy-analyze on texts.
/media on photos and videos.

---name: inboxdescription: Sorts client materials from inbox/. Verifies compliance with content-pipeline requirements (naming, formats, sizes). Produces a "ready / needs rework" report.---# /inboxFirst pass over client materials. **Does not convert** (that's `/media`), **does not analyze text** (that's `/copy-analyze`) — only sorts and validates.## When to use- Client says "materials uploaded".- After a new batch arrives (top-up).## Process**Step 1. Read `brief/08-content-pipeline.md`**Load requirements: naming, formats, sizes.**Step 2. Walk `inbox/` folders**For each folder (`texts/`, `media/photos/`, `media/videos/`, `brand/logo/`, `brand/fonts/`, `notes/`):- List files.- Verify: naming ASCII? correct format? size within limits?- Mark: ✅ OK / ⚠️ warning / ❌ reject.**Step 3. Specific checks****texts/:**- UTF-8?- Section headings present (`# Hero`, `# Offer`)?**media/photos/:**- Min side ≥ 2400px (read via image tool or `identify`).- sRGB (not Adobe RGB — otherwise color shift in the browser).- ASCII filenames.- No `IMG_1234.jpg`.**media/videos/:**- 1080p+.- Bitrate ≤ 20 Mbps.- H.264 in MP4 (not HEVC — poor browser support).**brand/logo/:**- SVG only.- Text outlined (check: is there a `<text>` tag inside SVG — if yes, NOT outlined).- No embedded raster (no `<image>` inside).**brand/fonts/:**- `.woff2` only.- `LICENSE.txt` present.**notes/:**- List with type (audio / image / md / other).- Transcribe voice notes (if possible).**Step 4. Produce the report**Create/update `inbox/inbox-report.md`:```markdown# Inbox report — 2026-08-08## Overall- Files total: N- OK: N | Warning: N | Reject: N## texts/- `copy.md` — ✅ OK (UTF-8, 5 sections)- `about.docx` — ⚠️ convert to MD## media/photos/- `hero-01.jpg` — ✅ 3600×2400, sRGB- `IMG_1234.jpg` — ❌ rename to ASCII- `founder.tif` — ⚠️ 6000×4000 > 6000px — will be cropped## brand/logo/- `logo.svg` — ❌ text not outlined (found <text>)## Summary for the client### Ready- Texts can go to /copy-analyze- Hero photo (after rename)### Needs rework- 3 photos: rename- Logo: outline (in Illustrator: Type → Create Outlines)- `about.docx`: convert to MD## Next- Wait for client rework- OR if user says "ship with what we have" — run /media on ok-files and /copy-analyze on texts```**Step 5. Extract links from notes/**If notes contain URLs — offer to move them into `references/` by aspect.## Rules- **Do not convert files here.** Only validate and sort.- **Do not complain if materials are thin.** Just report "we have X, for a full landing we need Y".- **Do not delete `reject` files.** Only mark in the report — user decides.## Next- `/copy-analyze` on texts.- `/media` on photos and videos.

Frequently asked questions

Inbox