Skip to content

SKILL.md

Latest commit

 

History

History
54 lines (41 loc) · 1.91 KB

SKILL.md

File metadata and controls

54 lines (41 loc) · 1.91 KB
name beeper-linkedin
description Detect and filter LinkedIn spam/prospection messages via Beeper MCP. Scans LinkedIn DMs, analyzes patterns, suggests responses with confirmation workflow. Use when asked to filter LinkedIn messages, detect prospection, or manage LinkedIn DMs via Beeper.
homepage https://github.com/manthis/linkedin-spam-filter
metadata
openclaw
emoji requires
💼
bins
mcporter
python3

openclaw-skill-linkedin-spam-filter

Detect LinkedIn spam/prospection messages via Beeper MCP. Generates suggested responses with confirmation workflow.

Usage

# Check LinkedIn messages
python3 scripts/linkedin-spam-filter.py [--dry-run] [--json]

# Test detection on text
python3 scripts/linkedin-spam-filter.py --test-text "Hi, I have an exciting opportunity..."

Configuration (env vars)

Variable Required Default Description
BEEPER_SERVER beeper Beeper MCP server name
MCPORTER_CMD mcporter mcporter CLI path
LINKEDIN_ROOM_PATTERN linkedin Pattern to match LinkedIn rooms
SPAM_PATTERNS (built-in) Regex patterns for detection
RESPONSE_TEMPLATES (built-in) JSON templates for responses
LINKEDIN_STATE ~/.openclaw-linkedin-state.json State file
LINKEDIN_LOG ~/logs/linkedin-spam-filter.log Log file

Workflow

  1. Detect — Scans LinkedIn DMs via Beeper MCP
  2. Analyze — Matches against prospection patterns
  3. Suggest — Generates response (FR/EN auto-detect)
  4. Notify — Reports to agent for confirmation
  5. Send — After human confirmation (not auto-sent)

Dependencies

  • mcporter CLI with Beeper MCP configured
  • Python 3.8+

OpenClaw Integration

cron:
  - name: "LinkedIn Spam Filter"
    schedule: "0 */6 * * *"
    command: "python3 scripts/linkedin-spam-filter.py --json"