xcom-rs

name: xcom-rs description: Interact with X/Twitter via the xcom-rs CLI (Rust). Use for posting tweets, replies, threads, searching, reading timelines/mentions, liking, retweeting, bookmarks, media upload, and user lookups. Use this skill whenever the user wants to do anything on X/Twitter — posting, reading, searching, monitoring mentions, managing bookmarks, or looking up users and their tweets. version: 1.0.0 author: tumf license: MIT platforms: [linux, macos] prerequisites: commands: [xcom-rs] metadata: hermes: tags: [twitter, x, social-media, xcom-rs] homepage: https://github.com/tumf/xcom-rs

xcom-rs — X/Twitter CLI

Agent-friendly Rust CLI for X/Twitter with structured output, OAuth2 PKCE auth, cost tracking, and idempotency support.

Install

cargo install --git https://github.com/tumf/xcom-rs.git

Verify:

xcom-rs --help
xcom-rs doctor --output json

Authentication

xcom-rs uses OAuth2.0 PKCE (preferred) or OAuth1.0a 3-legged flow. No manual API key juggling needed.

xcom-rs auth login          # Interactive browser-based OAuth2 PKCE
xcom-rs auth status          # Check current auth state
xcom-rs auth logout          # Revoke and clear tokens

Alternatively, set XCOM_RS_BEARER_TOKEN env var for app-only (read) access.

Check health with:

xcom-rs doctor --output json

This shows auth status, granted scopes, budget storage, and can probe the API with --probe.

Output Formats

All commands support --output json|yaml|text. Use --output json when parsing results programmatically. All JSON responses follow a consistent envelope:

{"ok": true, "type": "<command>", "schemaVersion": 1, "data": { ... }}

On error:

{"ok": false, "error": {"code": "RATE_LIMIT_EXCEEDED", "message": "...", "retryable": true}}

Commands Reference

Tweets

# Show a tweet
xcom-rs tweets show <tweet_id> --output json

# Get conversation thread for a tweet
xcom-rs tweets conversation <tweet_id> --output json

# Reply to a tweet
xcom-rs tweets reply <tweet_id> "reply text" --output json

# Post a thread (sequential replies)
xcom-rs tweets thread "first" "second" "third" --output json

# Like / unlike
xcom-rs tweets like <tweet_id>
xcom-rs tweets unlike <tweet_id>

# Retweet / unretweet
xcom-rs tweets retweet <tweet_id>
xcom-rs tweets unretweet <tweet_id>

Search

# Search recent tweets
xcom-rs search recent "query" --limit 10 --output json

# Search users
xcom-rs search users "query" --limit 10 --output json

Timelines

# Home timeline
xcom-rs timeline home --limit 20 --output json

# User's tweets
xcom-rs timeline user <username> --limit 20 --output json

# Mentions
xcom-rs timeline mentions --limit 20 --output json

Bookmarks

xcom-rs bookmarks list --limit 20 --output json
xcom-rs bookmarks add <tweet_id>
xcom-rs bookmarks remove <tweet_id>

Media

# Upload media (returns media_id for attaching to tweets)
xcom-rs media upload <file_path> --output json

Introspection

# List all commands with metadata (risk level, cost info)
xcom-rs commands --output json

# Get JSON schema for a command's input/output
xcom-rs schema <command_name> --output json

# Get detailed help including exit codes and examples
xcom-rs help <command_name> --output json

Idempotency

Write commands support --client-request-id <id> (or --client-request-id-prefix <prefix> for threads) to prevent duplicate posts on retry. Use --if-exists return|error to control behavior when a duplicate is detected.

xcom-rs tweets reply 123 "hello" --client-request-id my-reply-001 --if-exists return

Cost & Budget Controls

xcom-rs tracks API cost in credits. Control spending with:

--max-cost-credits <N>       # Fail if single operation exceeds N credits
--budget-daily-credits <N>   # Fail if daily total would exceed N credits
--dry-run                    # Estimate cost without executing

Budget data stored at ~/.local/share/xcom-rs/budget.json.

Non-Interactive Mode

For agent use, always pass --non-interactive to prevent prompts:

xcom-rs tweets reply 123 "hello" --non-interactive --output json

Exit Codes

Error Codes

Retryable: RATE_LIMIT_EXCEEDED, NETWORK_ERROR, SERVICE_UNAVAILABLE Not retryable: INVALID_ARGUMENT, MISSING_ARGUMENT, AUTHENTICATION_FAILED, AUTHORIZATION_FAILED

Agent Workflow

1. Check auth: xcom-rs auth status --output json
2. If not authenticated: xcom-rs auth login (needs interactive) or set XCOM_RS_BEARER_TOKEN
3. Use --output json --non-interactive for all commands
4. Parse the ok field to check success
5. For writes (reply, thread, retweet), confirm user intent first — these are public actions
6. Use --client-request-id for idempotent retries on write operations

Pitfalls

• Auth required: Most commands need OAuth2 login first. doctor will tell you what's missing.
• Rate limits: X API has per-endpoint rate limits. Retryable errors include RATE_LIMIT_EXCEEDED.
• Tweet posting: tweets reply and tweets thread are the write commands. There is no standalone tweet post — use tweets thread "single tweet" for a single post.
• Scopes: If writes fail, check doctor output for missing scopes (need tweet.write, like.write, etc.).
• Cost tracking: xcom-rs tracks credits locally. Set --budget-daily-credits to prevent runaway usage.