agentleFS
fliqrlabsfliqrlabs/instagram-creator-skills/skills/viral-engine/SKILL.md

viral-engine

Skill5 starsChanged 18 days agoMITView on GitHub

The full pipeline in one command: from an idea (or nothing) to a finished Instagram post with a scored hook, script or slides, caption, polish check, and a live comment-to-DM funnel, so the post earns reach and turns it into leads. Runs creator-brain, comment-mining, viral-hooks, reel-script or carousel, caption, polish and comment-to-dm in order. Use when the creator says "go viral", "make me a viral post", "do everything", "full post from this idea", "I want leads from Instagram", "what should I post today and set it up", or invokes the engine. Works offline, or with the Fliqr AI connector for live Instagram data.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 3 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Look for the creator profile.

    line 33
  2. Very specific:

    Check for the Fliqr AI connector (../instagram-playbook/connectors.md) and say which mode you're in: "Connected to @handle" or "Offline …

    line 34
  3. Very specific:

    Run viral-hooks: 6 formulas, scored, top 2 recommended.

    line 44
  4. Very specific:

    Run polish on every piece of text that will be published or sent.

    line 57
  5. Very specific:

    if the post isn't published yet, use is_next: true so the automation attaches to the next post

    line 62
  6. Very specific:

    if it's already live, use its media_id

    line 63
  7. Distinctive:

    Create it paused, show it, and activate it only on a yes.

    line 65
  8. Distinctive:

    Never fabricate proof to make a hook stronger.

    line 91
  9. Distinctive:

    If the creator only wants part of the pipeline, run that part.

    line 92

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

viral-engine

The other skills each do one job. This one runs the whole loop:

 FIND ──► HOOK ──► BUILD ──► CAPTION ──► POLISH ──► FUNNEL ──► (post) ──► MEASURE
  │                                                                          │
  └──────────────────────── next idea comes from what worked ◄───────────────┘

Viral reach with no funnel gives you views and $0. A funnel with no reach gives you nothing to convert. This engine always builds both.

0. Load context

  • Look for the creator profile. If there isn't one, run creator-brain first. With the Fliqr AI connector it takes about 2 minutes. Without it, ask for 3 captions and the 5 interview questions.
  • Check for the Fliqr AI connector (../instagram-playbook/connectors.md) and say which mode you're in: "Connected to @handle" or "Offline mode: paste data when I ask".

1. FIND (skip if the creator brought an idea)

Connected: run quick versions of growth-audit (top 3 outliers: what format and hook formula won) and comment-mining (top 3 audience questions). Offline: ask "What did people ask you about most this week?"

Propose 3 ideas, each with a goal (SEND / SAVE / LEAD), a format, and the evidence behind it. The creator picks one.

2. HOOK

Run viral-hooks: 6 formulas, scored, top 2 recommended. The creator picks, or you pick and say why.

3. BUILD

  • Reel → reel-script (timed script, on-screen track, shot list)
  • Carousel → carousel (slide plan + design notes)

4. CAPTION

caption: the keyword above the fold, ≤5 hashtags, the first comment.

5. POLISH

Run polish on every piece of text that will be published or sent. Nothing ships while it says FAIL.

6. FUNNEL

For LEAD posts, and for any post the creator expects to do well, run comment-to-dm. If the keyword has nothing real to deliver yet, run lead-magnet first.

  • if the post isn't published yet, use is_next: true so the automation attaches to the next post
  • if it's already live, use its media_id

Create it paused, show it, and activate it only on a yes.

7. LOG

Append the approved post to the post log (../instagram-playbook/post-log.md), with its formula, keyword and automation id.

8. HAND-OFF

One final block the creator can work from:

READY TO POST  ·  Reel 28s  ·  goal LEAD  ·  keyword EDIT
✓ hook   92  Time Collapse        ✓ script timed, loop ✓        ✓ polish PASS
✓ caption + first comment         ✓ funnel "EDIT → template" created (paused) ▸ activate?

Post it, then:
  · reply to the first 10 comments yourself in the first hour
  · share it to your story with a "reply EDIT" frame (story-funnel)
  · ask me "growth-audit 7d" next week. I'll check views against your median and triggers against comments

With a scheduling connector (posts.schedule), offer to queue the post instead of the manual "post it" step.

Rules

  • Each step's gate still applies: no automation goes live and nothing gets published without a yes.
  • Never fabricate proof to make a hook stronger. Ask for the real number or use a placeholder.
  • If the creator only wants part of the pipeline, run that part. The engine isn't a cage.
<!-- brand:footer -->

Part of Instagram Creator Skills by Fliqr AI. Live Instagram data and comment-to-DM automations come from the Fliqr AI connector (fliqr.ai).

<!-- /brand:footer -->

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.