AgentSkillsCN

linkt-signals

从 Linkt 中提取最新信号,并为拓展联系人展示相关信息。适用于用户希望查看商业信号、寻找待联系的潜在客户,或为 LinkedIn 拓展做好准备时使用。

SKILL.md
--- frontmatter
name: linkt-signals
description: Pull recent signals from Linkt and display contacts for outreach. Use when user wants to see business signals, find leads to contact, or prepare for LinkedIn outreach.
user-invocable: true
allowed-tools: mcp__linkt__list_signals_v1_signal_get, mcp__linkt__get_signal_v1_signal, mcp__linkt__list_entities_v1_entity_get, mcp__linkt__get_entity_v1_entity, mcp__linkt__list_icps_v1_icp_get

Linkt Signals Skill

Display recent business signals with associated company and contact information for sales outreach.

Workflow

Step 1: Fetch Recent Signals

Use mcp__linkt__list_signals_v1_signal_get to fetch signals:

  • Default: last 30 days
  • Page size: 10-20 signals
  • Sort by: detected_at descending (most recent first)

If no signals found, inform the user they may need to:

  1. Set up signal monitoring first (see python/03_signals/signals_from_sheet.py)
  2. Wait for the monitoring run to complete
  3. Check if the ICP has companies to monitor

Step 2: Enrich with Entity Data

For each signal, get the associated entity (company) using mcp__linkt__get_entity_v1_entity with the signal's entity_id.

Key fields to extract from company entity:

  • data.name - Company name
  • data.website - Company website
  • data.industry - Industry
  • data.employee_count - Company size
  • data.description - Company description

Step 3: Get Contacts for Each Company

For companies with interesting signals, fetch associated contacts using mcp__linkt__list_entities_v1_entity_get:

  • Filter by entity_type: "person"
  • Filter by icp_id from the company's ICP

Key contact fields:

  • data.name or data.full_name - Contact name
  • data.title or data.job_title - Job title
  • data.linkedin_url or data.linkedin - LinkedIn profile URL
  • data.email - Email address (if available)

Step 4: Present Results

Format output as follows:

markdown
## Recent Signals (Last 30 Days)

### 1. [Company Name] - [Signal Type Display Name]
**Signal:** [Summary from signal]
**Strength:** [Strong/Moderate/Weak] | **Score:** [0.85] | **Detected:** [Date]
**Source:** [source_url if available]

**Company:** [Industry] | [Employee Count] employees
[Company description snippet]

**Contacts:**
- [Name] ([Title]) - [LinkedIn](url)
- [Name] ([Title]) - [LinkedIn](url)

---

### 2. [Next Company]...

Step 5: Offer Outreach

After displaying signals, ask the user:

Would you like to reach out to any of these contacts? Enter a number to select a contact, or type 'skip' to exit.

If user selects a contact, provide context for the /linkt-outreach skill:

  • Signal summary and type
  • Company name and context
  • Contact name, title, and LinkedIn URL

Signal Types

Common signal types to highlight:

  • AI Initiatives - AI projects, ML implementations
  • AI Thought Leadership - LinkedIn posts about AI
  • AI Job Postings - Hiring for AI roles
  • funding - Funding rounds
  • leadership_change - New executives
  • hiring_surge - Rapid hiring
  • product_launch - New products
  • partnership - Strategic partnerships
  • expansion - Geographic or market expansion
  • acquisition - M&A activity
  • layoff - Workforce reductions
  • award - Industry recognition
  • pivot - Strategic direction change
  • regulatory - Compliance/regulatory news
  • rfp - Request for proposals
  • contract_renewal - Contract renewals
  • infrastructure - Technology investments
  • compliance - Compliance updates

Signal Score

Each signal has a score field (0.0-1.0) indicating relevance/priority:

  • 0.8-1.0: High priority signals - strong buying signals
  • 0.5-0.79: Medium priority - worth monitoring
  • 0.0-0.49: Lower priority - informational

Display the score alongside strength for prioritization.

Error Handling

  • If API returns no signals: Suggest setting up monitoring
  • If entity fetch fails: Skip that signal, continue with others
  • If no contacts found: Note "No contacts enriched yet" for that company

Example Output

markdown
## Recent AI Signals (Last 30 Days)

Found 3 signals across 2 companies.

---

### 1. TechCorp Inc - AI Thought Leadership
**Signal:** CEO published LinkedIn article on "How AI is Transforming Enterprise Sales"
**Strength:** Strong | **Score:** 0.92 | **Detected:** Jan 28, 2026
**Source:** [LinkedIn Post](https://linkedin.com/posts/...)

**Company:** Enterprise Software | 500-1000 employees
Leading provider of sales automation tools for mid-market companies.

**Contacts:**
1. Sarah Chen (VP of Sales) - [LinkedIn](https://linkedin.com/in/sarahchen)
2. Mike Johnson (Head of Partnerships) - [LinkedIn](https://linkedin.com/in/mikej)

---

### 2. DataFlow Systems - AI Job Postings
**Signal:** Posted 5 AI/ML engineering roles in the past week
**Strength:** Strong | **Score:** 0.88 | **Detected:** Jan 27, 2026

**Company:** Data Infrastructure | 200-500 employees
Cloud-native data platform for real-time analytics.

**Contacts:**
3. Alex Rivera (CTO) - [LinkedIn](https://linkedin.com/in/alexrivera)
4. Jordan Lee (VP Engineering) - [LinkedIn](https://linkedin.com/in/jordanlee)

---

**Select a contact for outreach (1-4) or 'skip':**