agentleFS
matthewdonsemail-labmatthewdonsemail-lab/log/.claude/skills/log-discord-meme-reply/SKILL.md

log-discord-meme-reply

Skill63 starsChanged yesterdayMITView on GitHub

Replying with a meme in a Discord chat. A log play (docs/useCases/discord/memeReply.md): Answer a moment in a Discord group or DM the way people there already do: with a meme. The chat is read first, [Memelord](https://www.memelord.com/sign-up?ref=kfwut9wc) finds a template that fits what was just said, the meme is gated like any other send, and it goes out as its media url, which Discord shows inline. Only Memelord's free tools are used (rule 41). The extension itself is described in [docs/extensions/memelord.md](../../extensions/memelord.md).

The file

Files
SKILL.md

Replying with a meme in a Discord chat

This is the log play discord/memeReply. Before anything else, call the log MCP tool getGuide({ useCase: "discord/memeReply" }): it returns this page and makes it the session's play, so every log tool's answer says which step you are on and a send with no play loaded is refused (rules 157, 158). Then follow the steps in order.

Draft every message from that person's own thread and matt's real replies (voice with no text, search), never from memory, and never send two that read alike (sameAsSent).

Answer a moment in a Discord group or DM the way people there already do: with a meme. The chat is read first, Memelord finds a template that fits what was just said, the meme is gated like any other send, and it goes out as its media url, which Discord shows inline. Only Memelord's free tools are used (rule 41). The extension itself is described in docs/extensions/memelord.md.

How the owner asks for it: "respond with memes in the <chat>", "drop a meme in <chat> about that", "read <chat> and meme it"

Run it from a fresh context

Say: "go on docs/useCases/discord/memeReply.md for <chat>". The agent reads this page and runs the steps.

The steps

#StepToolDone when
1Read the chat, and note its id from the first linediscordRead { account, channel: "<chat as the owner says it>", limit: 25 }the latest messages are read and the chat id is known
2Pick the moment: the last thing said that a person would answer with a reactionthe agent, from the chatone message, named with its time and author
3Search for it in a few words; add keyword matching or a type when neededmemelordSearch { query: ["<the moment>"], method: "combined", limit: 8 }a template whose name and description fit the moment
4When nothing fits, look at what is trending, each with what it meansmemelordTrending { limit: 8 }a template that fits, or no meme is sent
5Gate it against how the owner writes in that chatvoice { to: "<chat>", with: "<chat>", context: "<the moment and the meme>", text: "<media url>" }PASS
6Send it by the chat id, the media url on a line of its owndiscordSendLive { account, channel: "<chat id>", text: "<media url>", approvedBy: "<owner>", confirmLive: true }Discord answers sent with the message id

What the runs taught

  • The owner's taste ranks the results. Search leads with the owner's taste and ranks each template by it (deep fried, larping, streamer reacts like FlightReacts, Zias and Bloo, shitposts, and the memes the owner has sent): pick from the top. On "waiting" it put Skeleton Underwater first and Sad Kermit last.
  • Short queries find more. "deciding how much money to charge a client" found one weak clip; "money" with method: "combined" found eight still memes (2026-11-02, the log group).
  • Send by the chat id. A chat named in words makes the engine list every chat first, which is one more call, and the account's governor then waits a random gap before the send (about 30 s on 2026-11-02). With the id the send is one call.
  • A meme can joke about money; it never names a price. On 2026-11-02 the owner asked the log group "how much shall i charge", and the answer sent was Floyd Mayweather's Cash Stacks, not a number.
  • One meme, then wait. The chat answers back; never send a run of memes (rule 94).

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.