X Research

General-purpose agentic research over X/Twitter. Decompose any research question into targeted searches, iteratively refine, follow threads, deep-dive linked content, and synthesize into a sourced briefing.

Uses the bird CLI for all X/Twitter data access.

Prerequisites

bird must be installed and authenticated:

# Install
brew install steipete/tap/bird
# or: npm install -g @steipete/bird

# Verify auth
bird check
bird whoami

If query IDs go stale (404 errors): bird query-ids --fresh

CLI Reference

Search

bird search "<query>" -n <limit>
bird search "<query>" --all --max-pages 3

Search operators (used inside the query string):

• from:username — tweets from a specific user
• -is:retweet — exclude retweets
• -is:reply — exclude replies
• has:links — only tweets with links
• url:github.com — tweets linking to a domain
• OR — combine terms: (opus OR claude)
• -keyword — exclude keyword
• min_faves:N — minimum likes
• min_retweets:N — minimum retweets

Examples:

bird search "BNKR" -n 10
bird search "from:frankdegods" -n 20
bird search "(opus 4.6 OR claude) trading" --max-pages 2
bird search "$BNKR (revenue OR fees) min_faves:5"
bird search "AI agents -is:retweet -is:reply has:links" -n 15
bird search "from:steipete" --all --max-pages 3

Profile / User Tweets

bird user-tweets @handle -n 20        # User's recent tweets
bird about @handle                     # Account origin/location info

Thread

bird thread <url-or-id>               # Full conversation thread

Single Tweet

bird read <url-or-id>                 # Read a single tweet
bird <url-or-id>                      # Shorthand

Replies

bird replies <url-or-id>              # List replies to a tweet
bird replies <id> --all --delay 1000  # Paginate all replies

Mentions

bird mentions                         # Tweets mentioning you
bird mentions --user @handle          # Mentions of another user

Output Modes

--json          # JSON output (useful for programmatic processing)
--plain         # No emoji, no color (script-friendly)

Research Loop (Agentic)

When doing deep research (not just a quick search), follow this loop:

1. Decompose the Question into Queries

Turn the research question into 3-5 keyword queries using X search operators:

• Core query: Direct keywords for the topic
• Expert voices: from: specific known experts
• Pain points: Keywords like (broken OR bug OR issue OR migration)
• Positive signal: Keywords like (shipped OR love OR fast OR benchmark)
• Links: url:github.com or url: specific domains
• Noise reduction: Add -is:retweet -is:reply for cleaner results
• Crypto spam: Add -airdrop -giveaway -whitelist if crypto topics flooding

2. Search and Extract

Run each query via bird search. After each, assess:

• Signal or noise? Adjust operators.
• Key voices worth searching from: specifically?
• Threads worth following via bird thread?
• Linked resources worth deep-diving with web_fetch?

3. Follow Threads

When a tweet has high engagement or is a thread starter:

bird thread <url-or-id>

4. Deep-Dive Linked Content

When tweets link to GitHub repos, blog posts, or docs, fetch with web_fetch. Prioritize links that:

• Multiple tweets reference
• Come from high-engagement tweets
• Point to technical resources directly relevant to the question

5. Synthesize

Group findings by theme, not by query:

### [Theme/Finding Title]

[1-2 sentence summary]

- @username: "[key quote]" (NL, NI) [Tweet](url)
- @username2: "[another perspective]" (NL, NI) [Tweet](url)

Resources shared:
- [Resource title](url) -- [what it is]

6. Save

Save research output to ~/clawd/drafts/x-research-{topic-slug}-{YYYY-MM-DD}.md.

Refinement Heuristics

• Too much noise? Add -is:reply, use min_faves:N, narrow keywords
• Too few results? Broaden with OR, remove restrictive operators, increase --max-pages
• Crypto spam? Add -$ -airdrop -giveaway -whitelist
• Expert takes only? Use from: or min_faves:50
• Substance over hot takes? Search with has:links

Watchlist (Manual)

Maintain a list of key accounts to periodically check:

bird user-tweets @account1 -n 5
bird user-tweets @account2 -n 5

Store your watchlist in data/watchlist.md and run through it when doing periodic checks.

File Structure

x-research-skill/
├── SKILL.md           (this file)
└── data/
    └── watchlist.md   (accounts to monitor)