agentleFS
hash-anmolhash-anmol/skills/skills/in-progress/video-publish-workflow/SKILL.md

video-publish-workflow

Skill2 starsChanged 5 days agoMITView on GitHub

In progress. Run the whole post-edit pipeline for a finished reel: lead magnet, hosting, cover, email, Buffer posts, Zorcha comment-to-DM automation. Use when the user says 'run the video workflow', 'publish this reel', 'post it and set up the automation', or asks for any step after the edit is done.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Ask these in one message, then work through the steps in order.

    line 14
  2. Very specific:

    Put the approved PDF on the user's Google Drive, set it to anyone with the link, and confirm a cookie-less download matches the local …

    line 34
  3. Distinctive:

    Ask for a final yes with the exact caption, title, channels and file shown.

    line 52

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Video Publish Workflow (in progress)

Take a finished, approved reel from "the edit is done" to "posted, with a comment-to-DM automation delivering the lead magnet". Every step that changes something outside this machine waits for an explicit yes, because posts and DMs cannot be unsent. Details that only some runs need live in references/.

This skill is beta. It was written from one real run, so expect gaps and fix them in place.

0. Ask everything first

Ask these in one message, then work through the steps in order. Take answers from the conversation where they already exist.

  1. The final video file path, and the reel's transcript or script.
  2. The exact caption and the YouTube title (use the user's words verbatim).
  3. Channels: Instagram, YouTube, or both. Post now or scheduled.
  4. What the lead magnet should be (the resource the reel promises), the page count limit (default 4 to 5), and the brand colours.
  5. The comment trigger: any comment, or one keyword. Whether follow-gating is wanted.
  6. Earlier reels whose commenters should also get the DM, and which older automations should be stopped afterwards.
  7. The email address for the cover (the user's own).

Done when: each answer is written down in one line, and anything missing has been asked rather than guessed.

1. Lead magnet

Build the PDF the reel promises, then stop for review. Follow references/lead-magnet.md for research, voice, layout and the layout measurement gate.

Done when: the user has approved the rendered PDF, and the page count and link annotations have been checked.

2. Host it

Put the approved PDF on the user's Google Drive, set it to anyone with the link, and confirm a cookie-less download matches the local SHA-256. Follow references/hosting-and-posting.md, "Hosting". Do not use anonymous upload sites; ask first if Drive is unavailable.

Done when: the public link is recorded and its downloaded bytes hash equal to the local file.

3. Cover

Make the Instagram cover with the user's thumbnail agent, giving it a smiling reference frame cut from the raw footage at full resolution, the exact copy, and any game or product art as a design-only asset. Look at the result at full size, check the text letter by letter, and keep the version that stays legible at thumbnail size. See references/hosting-and-posting.md, "Cover".

Done when: one cover is chosen by the user and the exact file path is recorded.

4. Email the cover

Send the cover to the user's own address. If the Gmail connector can only draft, or reports the connection as invalid, say so and use the agent mail route instead of dropping the step. See references/hosting-and-posting.md, "Email".

Done when: a sent message is read back with both attachments present, or the user has been told exactly which step is blocked.

5. Post

Ask for a final yes with the exact caption, title, channels and file shown. Then post through Buffer using the public video link. Check status until it says sent, and record each permalink. See references/hosting-and-posting.md, "Posting".

Done when: every requested channel shows status sent with a live link, or the failure is reported verbatim.

6. Comment-to-DM automation

Create the automation in the user's browser through the Chrome extension, not through the Zorcha MCP unless its workspace has access. Follow references/zorcha-browser.md. Duplicate the newest follow-gated automation, pin the exact reel by link, replace every button link with the hosted guide, go live, and verify from the list.

Done when: the new automation is Live, pinned to the reel's permalink, and both the guide button and the follow-up button open the hosted guide.

7. Earlier reels and cleanup (only if asked in step 0)

Duplicate the new automation once per earlier reel. Before each rerun, show the count of waiting comments and get a yes, because a rerun sends real DMs. Stop older automations only after the user has confirmed the list, in small batches, and re-check the list afterwards. See references/zorcha-browser.md, "Reruns and stopping".

Done when: each earlier reel has a Live automation and a started rerun, and the list shows the intended automations Live and the rest stopped.

8. Report

Give the user the public guide link, each post link, the automation names and states, what was not done, and anything unverified.

Done when: the report names the link, posts, automations, and open items.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.