agentleFS
ramkesh430ramkesh430/facebook-agent-skill/skills/fb-boost/SKILL.md

fb-boost

Skill1 starsChanged 16 days agoMITView on GitHub

Decide whether a Facebook post is worth boosting, then write the boost or ad - which post, goal, audience, budget sanity check, primary text and headline within the truncation limits - without ever spending money. Use when the user asks "should I boost this post", "write a Facebook ad", "how much should I spend", "who should I target", or "my boost didn't work".

The file

We split this skill into the rules it gives an agent

10 rules. About 0 are common, 9 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    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:

    Not financial advice. It's a checklist for spending a small budget carefully.

    line 18
  2. Distinctive:

    Don't boost a post to "fix" weak reach.

    line 29
  3. Distinctive:

    If the user doesn't know, it's almost always Messages for a small local business.

    line 44
  4. Distinctive:

    Age range: only if the product truly excludes ages.

    line 50
  5. Distinctive:

    Never target or exclude by sensitive traits (health, religion, ethnicity, etc.).

    line 55
  6. Distinctive:

    Start small and fixed: a few days, a daily amount the user can lose without stress.

    line 61
  7. Distinctive:

    Stop or change after 3-4 days if it's not hitting that.

    line 65
  8. Distinctive:

    If the boost is an ad with new text, write it and lint it:

    line 69
  9. Distinctive:

    No bait, no fake urgency, no before/after claims that break Meta's ad policies (health, weight loss, finance claims need extra care).

    line 79
  10. Very specific:

    Run it through /fb-human.

    line 82

≈ “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

fb-boost

Boosting is the cheapest paid lever on Facebook and the easiest to waste money on. This skill makes the decision and writes the copy. It never spends, never logs in to Ads Manager, never enters payment details. The user sets up and pays for the boost themselves.

Not financial advice. It's a checklist for spending a small budget carefully.

Step 1: should this be boosted at all?

Boost a post only if it already worked organically. Run /fb-audit or ask for the post's numbers. A good candidate:

  • did 2x+ the Page's median reach, or got real shares and messages
  • is still true and in stock (no expired dates or prices)
  • has one clear action

Don't boost a post to "fix" weak reach. Paid reach on a post people ignore is paying to be ignored.

Step 2: one goal

Pick the goal Facebook offers that matches the one action:

user wantsgoal to choose
people to messageMessages (Messenger / WhatsApp)
visits to a websiteLink clicks / traffic
more people to see a ReelEngagement or video views
callsCalls
local foot trafficthe local-awareness / store-visit option, if offered

If the user doesn't know, it's almost always Messages for a small local business.

Step 3: audience

  • Location first. Radius around the shop, or the cities served. Most wasted boosts are shown to the wrong country.
  • Age range only if the product truly excludes ages.
  • Interests: few and specific, or none; Facebook's own delivery is often better than a stack of guessed interests.
  • Retargeting beats strangers: people who engaged with the Page or watched its videos, when the user has enough of them.
  • Never target or exclude by sensitive traits (health, religion, ethnicity, etc.). Meta restricts this, and housing, jobs, credit and politics have special rules. Flag it if the post touches those.

Step 4: budget sanity check

  • Start small and fixed: a few days, a daily amount the user can lose without stress. They name the number, not you.
  • Decide before spending what result counts as "worked" (e.g. cost per message under {{their number}}).
  • Stop or change after 3-4 days if it's not hitting that.

Step 5: the copy

If the boost is an ad with new text, write it and lint it:

python3 ../fb-post/post.py ad.txt --mobile 125
  • Primary text: the first ~125 characters are what show before "See more" in most ad placements. The offer and the "for whom" go there.
  • Headline: ~40 characters. The benefit, not the brand name.
  • One action, matched to the goal.
  • No bait, no fake urgency, no before/after claims that break Meta's ad policies (health, weight loss, finance claims need extra care).

Run it through /fb-human.

Output

BOOST PLAN
post:       "If you live in Patan..." Reel - 6.2x median, 9 messages organic
goal:       Messages
audience:   10 km around Patan, 22-55, no interests
budget:     {{daily amount}} x 4 days, set by you
success:    cost per message under {{your number}}
copy:       primary 118/125 chars · headline 31/40 · linter READY

The user creates and pays for the boost.

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.