Find intent signals. Qualify leads. Draft outreach.
Open-source CLI that collects tweets, judges relevance via LLM, scores leads, and generates human-approved replies.
From signal to sent — in one pipeline
Seven stages, fully configurable, human-approved at every step.
Collect
Search X for matching tweets
Normalize
Clean text, extract entities
Judge
LLM classifies relevance
Score
Weighted 0-100 lead ranking
Draft
AI generates reply drafts
Approve
Human reviews every draft
Send
Rate-limited delivery
Built for founders and growth teams
Everything you need to find high-intent leads and engage them intelligently.
LLM-Powered Judging
Claude or GPT classifies tweet relevance with structured confidence scores and reasoning.
Weighted Lead Scoring
0-100 score combining relevance, author authority, engagement, recency, and intent strength.
AI Draft Generation
Context-aware reply drafts using your project persona, tone, and voice guidelines.
Human-in-the-Loop
Approve, edit, or reject every single draft before it's sent. No auto-pilot.
Rate-Limited Sending
Configurable hourly, daily, and monthly caps with jitter. Compliance-first design.
Outcome Tracking
Monitor if your replies get liked, replied to, or lead to follows and conversions.
See it in action
Real CLI output from the Syntrix pipeline.
Get started in 60 seconds
Three steps from zero to your first pipeline run.
Install
One command via PyPI.
pip install signalopsConfigure
Set your API keys.
export ANTHROPIC_API_KEY=sk-...
export X_BEARER_TOKEN=AAAA...Run
Execute the full pipeline.
signalops project set spectra
signalops run all --dry-run