Social Cli logo

Social Cli

Organization
letta-ai
social-cli

Agent-optimized CLI for Bluesky (ATProto) and X (Twitter). YAML in, YAML out, exit codes for automation. Use when the task involves posting, replying, reading feeds, searching, annotating URLs, or running a sync/check/dispatch agent loop across social platforms.

Overview

Publisherletta-ai
Repositoryskills
Skill namesocial-cli
Stars
144
Forks
25
Bundled files
5
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 5 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by letta-ai on GitHub. Read the source before you install it.

Installation

Install the Social Cli AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/letta-ai/skills.git /tmp/skills
mkdir -p .claude/skills
cp -r /tmp/skills/tools/social-cli .claude/skills/social-cli
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Social Cli in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Social Cli on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Social Cli is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

social-cli

Agent-optimized social CLI. Bluesky + X. YAML in, YAML out, exit codes for automation.

Repo: https://github.com/letta-ai/social-cli

Setup

1. Install

bash
git clone https://github.com/letta-ai/social-cli.git
cd social-cli
pnpm install
pnpm build

Link the binary globally or invoke via node dist/cli.js / pnpm start from the repo directory.

2. Credentials

Create a .env in the working directory where you run social-cli:

bash
# Bluesky / ATProto
ATPROTO_HANDLE=you.bsky.social
ATPROTO_APP_PASSWORD=xxxx-xxxx-xxxx-xxxx
ATPROTO_PDS=https://bsky.social          # optional, defaults to bsky.social

# X / Twitter (OAuth 1.0a)
X_API_KEY=...
X_API_SECRET=...
X_ACCESS_TOKEN=...
X_ACCESS_TOKEN_SECRET=...
X_BEARER_TOKEN=...                       # optional, for app-only endpoints

Only credentials for platforms you actually use are required.

3. Verify

bash
social-cli whoami
social-cli rate-limits

Agent loop (sync → check → dispatch)

The canonical automation workflow. Each phase reads/writes YAML files in the working directory.

bash
social-cli sync                    # pull notifications → inbox-{platform}.yaml
social-cli check || exit 0         # anything actionable? exit 1 = nothing to do
# read inbox, decide, write outbox-{platform}.yaml
social-cli dispatch                # execute outbox, archive results
  • sync: deduplicates against sent_ledger-{platform}.yaml and caps inbox size.
  • check: exits 0 if inbox has actionable items, 1 otherwise. Wrap with || exit 0 in cron loops.
  • dispatch: executes the outbox atomically. Writes dispatch_result-{platform}.yaml and appends to sent_ledger-{platform}.yaml for replay protection.

Quick commands

bash
# Post / reply / thread
social-cli post "Hello world" -p bsky
social-cli reply "Thanks" --id "at://did:plc:.../app.bsky.feed.post/abc" -p bsky
social-cli thread "p1" "p2" "p3" -p bsky

# Engagement
social-cli like "at://..." -p bsky
social-cli delete "at://..." -p bsky
social-cli follow "handle.bsky.social" -p bsky

# Reading
social-cli search "query" -p bsky -n 10              # → stdout YAML
social-cli feed -p bsky -n 20                        # → feed.yaml (or -o - for stdout)
social-cli feed --feed "at://did:.../app.bsky.feed.generator/name" -n 10  # custom feed
social-cli profile "handle.bsky.social" -p bsky
social-cli whoami
social-cli rate-limits

Annotations (Bluesky only)

Uses the at.margin.note lexicon (W3C Web Annotation model). Annotations work on any URL, not just ATProto posts. They appear in margin.at and Semble.

bash
# Annotate a web page
social-cli annotate "Note about this article" --target https://example.com

# Anchor to an exact passage
social-cli annotate "Key insight" --target https://example.com \
  --quote "exact passage from the page" --motivation highlighting

Motivations: commenting, highlighting, questioning, describing, linking.

Inbox format (inbox-{platform}.yaml)

yaml
notifications:
  - id: "at://did:plc:xxx/app.bsky.feed.post/abc"
    platform: bsky
    type: mention          # mention, reply, like, follow, repost, quote
    author: someone.bsky.social
    authorId: "did:plc:xxx"
    postId: "at://..."
    text: "Hey, what do you think?"
    timestamp: "2026-03-25T12:00:00Z"
    parentPostId: "at://..."     # for replies
    parentPostText: "..."        # context
    rootPostId: "at://..."       # thread root
    rootPostText: "..."

Outbox format (outbox-{platform}.yaml)

Write decisions as a dispatch list. Each entry is a single action.

yaml
dispatch:
  - reply:
      platform: bsky
      id: "at://did:plc:xxx/app.bsky.feed.post/abc"
      text: "Thanks for the mention"

  - post:
      text: "Hello from my agent"
      platforms: [bsky, x]          # post to both

  - thread:
      platform: bsky
      posts:
        - "Thread post 1"
        - "Thread post 2"

  - like:
      platform: bsky
      id: "at://..."

  - annotate:
      platform: bsky
      id: "https://example.com/article"
      text: "Key observation"
      motivation: commenting
      quote: "exact text to anchor to"

  - ignore:
      id: "notif_003"
      reason: "spam"

Dispatch results and exit codes

dispatch_result-{platform}.yaml is written after every dispatch. Exit codes:

CodeMeaning
0All actions succeeded
1Invalid outbox (schema error, missing creds)
2Partial failure (some succeeded, some failed)

Thread failures include a resumeFrom field with the index and remaining posts so you can retry just the tail.

Character limits

PlatformLimit
Bluesky300 chars
X280 chars

The CLI rejects over-limit posts before hitting the API.

Platform differences

FeatureBlueskyX
Annotations (at.margin.note)YesNo
SearchYesYes
FeedYes (custom feeds)Yes (home/user)
ThreadsYesYes
Notificationsmention, reply, like, follow, repost, quotementions only
Quote post contextYesYes

Resilience

  • Retries 3x with exponential backoff on 429s, 5xx, and network errors. Respects Retry-After.
  • Bluesky session auto-refreshes on token expiry.
  • Atomic file writes (tmp + rename) — no partial inbox/outbox corruption.
  • Thread resume on partial failure via resumeFrom.
  • sent_ledger-{platform}.yaml prevents duplicate dispatch across runs.

Working directory layout

Everything is scoped to your current working directory, so you can run multiple agents with isolated state:

./
├── .env
├── inbox-bsky.yaml          # sync output
├── inbox-x.yaml
├── outbox-bsky.yaml         # your decisions
├── outbox-x.yaml
├── dispatch_result-bsky.yaml  # last dispatch outcome
├── dispatch_result-x.yaml
├── sent_ledger-bsky.yaml      # replay protection
├── sent_ledger-x.yaml
└── feed.yaml                  # optional feed snapshots

References

  • references/commands.md: full command map with all flags
  • references/outbox-schema.md: complete outbox YAML schema
  • references/agent-loop.md: patterns for cron/systemd automation loops

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Social Cli AI skill do?

Agent-optimized CLI for Bluesky (ATProto) and X (Twitter). YAML in, YAML out, exit codes for automation. Use when the task involves posting, replying, reading feeds, searching, annotating URLs, or running a sync/check/dispatch agent loop across social platforms.

Why use Social Cli on TypingMind?

Because you install it once and use it with any model. Social Cli is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Social Cli in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/letta-ai/skills/tree/main/tools/social-cli. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Social Cli?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Social Cli?

As many as you like. As long as a model supports skills, you can use Social Cli with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Social Cli AI skill free?

It is published on GitHub by letta-ai. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇