agentleFS
JakeschincariolJakeschincariol/x-agent-skill/skills/x-comment/SKILL.md

x-comment

Skill4 starsChanged 45 days agoMITView on GitHub

Write replies on other people's X posts and threads that read as a person with something to add, not a reply guy. Use when the user pastes a post and wants a reply, says "reply to this", "comment on this thread", "engage with this", "what do I say here", "quote tweet this", or wants a batch of replies for their daily engagement round.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 1

    1. swayamdhawan/x-agent★ 0100% same

    Edited · 0

    None found.

    Wrote their own version · 0

    None found.

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository10 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Reply: when you are adding to their conversation.

    line 29
  2. Distinctive:

    Quote: when you have a take for your own audience that needs their post as the context.

    line 31
  3. Distinctive:

    Say the specific thing: If the reply could sit under any post on the topic, it is noise.

    line 59
  4. Distinctive:

    Do not post the reply. Do not drive a browser to publish it.

    line 99

≈ “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

x-comment

On X, replying to other people is how small accounts grow. A good reply under a post with 300,000 views is seen by more people than most of your own posts, and the people reading it already care about the topic. A bad one is seen by the same people, which is the problem.

Input

The user pastes the post (and the author's handle and what they do, if they have it). Screenshots are fine. If they give a URL you cannot open, ask them to paste the text. Do not guess what a post said, and do not use a browser to scrape anyone's timeline.

Reply or quote

Decide first, and say which:

  • Reply when you are adding to their conversation. Your reply lives in their thread, in front of their audience.
  • Quote when you have a take for your own audience that needs their post as the context. A quote is your post. It needs to stand on its own.
  • Never quote to dunk on a smaller account. It reads exactly the way it is, and a mute is weighted heavily against you in X's own ranker.

The nine reply types

Pick by what the post actually is. Never default to type 1.

#typewhenshape
1Add a datuma claim you can support with a number"Same here: 40% of our..."
2Add the missing caseright, but incomplete"Holds until {condition}. Then..."
3Respectful disagreeyou think it is wrongthe agreement first, then the fork
4Extend one lineone sentence in the post is the good onebuild on that sentence
5Ask the real questionthe post skipped the hard partone question, specific
6The receiptyou have done the thingwhat happened, in two sentences
7The correctiona factual errorbe right, be brief, be kind, be sure
8The reframeright facts, wrong frame"Another way to read this:"
9The one-linerthe post needs nothing, you want presenceunder 12 words, funny or true

Rules

  • One to three sentences. Most good replies are under 200 characters.
  • Never open with "Great post", "This 👆", "So true", "Facts", "💯", "Bookmarked", "Couldn't agree more", or the author's name and an exclamation mark. All invisible.
  • Never restate the post.
  • Say the specific thing. If the reply could sit under any post on the topic, it is noise.
  • Early beats clever. A reply in the first 15 minutes of a big account's post is seen. A better one two hours later is not. That is why this skill batches.
  • Replies between people who follow each other get a boost in X's published ranker (run python3 ../x-plan/weights.py). Peers matter.
  • Premium replies are ranked higher. X says so and calls the effect "slight". Mention it only if the user asks; it is a purchase decision, not advice.

Output

Two options of different types, labelled, plus one line on which to post. Both through /x-human first: a reply with an em dash is more obviously machine-written than a post, because replies are short and read closely.

REPLY OPTIONS  (on @author's post about pricing)

[6 · Receipt]
We raised prices 40% in March and lost 2 of 31 clients. The two we lost
were the two we argued with most.

[5 · Real question]
What happened to churn on the annual plans? That is where it bit us.

Post the receipt. It is your own number and it concedes a cost, which is
what people reply to.

Batch mode

For an engagement round, ask for 5 to 15 posts pasted in one message and return one reply each, in one block. Keep a running note in ~/.claude/x/log.md of who the user replied to this week. Replying to the same three people every day is visible, and it looks like what it is.

Never

Do not post the reply. Do not drive a browser to publish it. X's automation rules prohibit automated replies and mentions, including replies triggered by keyword searches, and AI reply bots need X's written approval. This skill writes the reply. The user posts it by hand.

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.