Skip to content

Skills / video-publish

video-publish

Part of the video library (loaded by /video, not run directly): a per-platform publish pack for a checked video; never uploads and refuses a video that failed QA.

OfficialRuns scriptsAuditB
Latest
v0.1.2
Aisle
video
Works with
Claude Code
Stars
2

Install

video-publish is a part of the video library. You do not run it yourself: you type /video and describe what you want, and Claude uses this part when it is needed. Installing the library installs this part too.

Install video

npx @agentsaisle/cli add video

Runs scriptsThis skill includes scripts

Your agent may run these scripts on your machine. A maintainer has read every one in full, and the CLI lists the files when you install. You can read them on GitHub first.

Read the files on GitHub

Audit

Checked with aisle audit (version 0.1.2). Run it yourself: aisle audit video-publish

B
Critical
0
High
0
Medium
1
Low
0

11 files · audited 2026-11-03 · rules 1.0.0

MediumAA-P02Can write files and run interpreters without disable-model-invocation

SKILL.md:5

Read Write Glob Bash(node *)

Why The agent may load this skill on its own and then write files and run code in one go.

Fix Set disable-model-invocation: true so the user starts it, or narrow the Bash patterns.

What this skill can use

Tools

  • Read
  • Write
  • Glob
  • Bash(node *)

Scripts

  • scripts/_lib.mjs
  • scripts/copycheck.mjs
  • scripts/ffwrap.mjs
  • scripts/publish.mjs
  • scripts/run.mjs

Domains mentioned

None

Automated checks found the patterns listed here. An audit reduces risk; it is not a guarantee of safety. How audits work · Read the rules

Show this grade in a README

The badge updates when the registry is rebuilt. Paste this Markdown into the README:

[![aisle audit](https://img.shields.io/endpoint?url=https%3A%2F%2Fwww.agentsaisle.dev%2Fbadges%2Fvideo-publish.json)](https://www.agentsaisle.dev/skills/video-publish)

SKILL.md

Rendered from commit 2ee11a4

Video publish

Turns a finished render into something ready to post. It prepares files and checks text; it never uploads, logs in or contacts a platform. <skill> below is the folder that holds this SKILL.md.

Use when

  • A project has a final render in out/, a passing video-qa report and credits.json, and the user wants it ready for one or more platforms.
  • A checked video is ready to prepare for --platforms youtube,shorts,reels,tiktok,linkedin.

Don't use when

  • The render is missing or QA has not passed. Name the step to run; do not work around the gate. The user may override a failed QA in words with a reason; use --override-qa "<reason>", never decide for them.
  • The user wants the video uploaded. Say this skill cannot, and the pack is for uploading by hand.
  • The video is the wrong shape for the platform (a 16:9 video for Reels). Say so; render it again in the platform's format (meta.format).

Steps

  1. Read brief.json, script.md and facts.md (once). Write publish/copy.json with title, description and hashtags (references/copy.md). Use only words and numbers from the script and facts.
  2. Run node <skill>/scripts/publish.mjs <project> --platforms <list> (add --no-card to skip the designed thumbnail). It checks the QA gate, the shape and length per platform and your text, and writes nothing if anything fails.
  3. If it printed REFUSED lines, fix each (the text, the shape, a missing QA or credits) and run it again. Do not edit platforms.json to make a limit pass.
  4. Show the pack: platforms, files, the kids answer, and any note. Read exports/<platform>/disclosure.md and tell the user which questions they must answer on the platform.
  5. Remind the user to upload by hand and that platform rules change (references/platforms.md).

Budget

  • Search first (grep, glob, symbol search). Read line ranges, not whole files.

  • Read at most 8 files per step. If you need more, write down what you learned and narrow the question.

  • Do not re-read a file you just wrote or edited.

  • Prefer the smallest command output that answers the question: head, tail, quiet flags, filtered lists.

  • Run the script once per set of changes. Do not open the encoded video or thumbnails to check them; the manifest has hashes and sizes.

Output

  • The platforms written, the number of files each, and the kids answer (yes, no or ask).
  • Every note the script printed (a skipped card, no chapters) and the override, if one was used.
  • That nothing was uploaded.

Done when

  • The script exited 0, exports/publish.json exists with "uploaded": false, and the user was told what to answer on each platform.
  • No figure in the text came from anywhere but the script or facts, and nothing was sent anywhere.

Version history

VersionCommitInstall
0.1.2latest2ee11a4add video-publish

Report this skillAgent Aisle on GitHub