AGENT SKILLBY CLAUDE-OFFICE-SKILLS
Lead routing
AT A GLANCE
Lead routing is a go-to-market skill for AI agents. Intelligent lead assignment and routing - AI-powered scoring, territory mapping, round-robin distribution, and workload balancing. The claude-office-skills/skills repository on GitHub, maintained by claude-office-skills, has 414 stars and the MIT license. To add it, run npx skills add claude-office-skills/skills --skill lead-routing
- SOURCE REPOSITORY
- claude-office-skills/skills (opens GitHub in a new tab)
- REPO STARS
- SOURCE REPOSITORY
- LICENSE
- MIT
- CATEGORIES
- SKILL.MD SIZE
- 12 KB510 LINES
QUICK ANSWER
- WHAT IT IS01
- Intelligent lead assignment and routing - AI-powered scoring, territory mapping, round-robin distribution, and workload balancing. Its edge is a particular angle on paid media, giving the agent tighter constraints than a plain lead routing request.
- INPUTS02
- Context for paid media: your goals, audience, constraints, and any source material the skill asks for.
- OUTPUT03
- A ready-to-use result for paid media: the analysis, copy, or recommendations the agent produces.
INSTALL
One command adds this skill from its source repository.
npx skills add claude-office-skills/skills --skill lead-routingREAD THE SKILL.MD BELOW BEFORE YOU RUN IT. SKILLS INSTRUCT YOUR AGENT.
SKILL.MD
The full file, as published in the source repository. Copy it, download it or open it on GitHub.
ON THIS PAGE · 15
Lead Routing#
Intelligent lead assignment and routing system with AI-powered scoring, territory mapping, round-robin distribution, and workload balancing. Based on n8n's HubSpot/Salesforce automation templates.
Overview#
This skill covers:
- Lead scoring and qualification
- Territory-based routing
- Round-robin distribution
- Workload balancing
- SLA monitoring and escalation
Routing Strategies#
1. Rule-Based Routing#
routing_rules:
# By Company Size
- name: "Enterprise Routing"
condition:
company_size: ">= 500"
OR:
annual_revenue: ">= $10M"
assign_to: "Enterprise Team"
priority: high
sla: 1_hour
- name: "Mid-Market Routing"
condition:
company_size: "100-499"
assign_to: "Mid-Market Team"
priority: medium
sla: 4_hours
- name: "SMB Routing"
condition:
company_size: "< 100"
assign_to: "SMB Team"
priority: standard
sla: 24_hours
# By Geography
- name: "APAC Routing"
condition:
country: ["China", "Japan", "Singapore", "Australia"]
assign_to: "APAC Team"
timezone_aware: true
- name: "EMEA Routing"
condition:
country: ["UK", "Germany", "France", "Netherlands"]
assign_to: "EMEA Team"
- name: "Americas Routing"
condition:
country: ["US", "Canada", "Brazil", "Mexico"]
assign_to: "Americas Team"
# By Industry
- name: "Healthcare Specialist"
condition:
industry: ["Healthcare", "Pharmaceuticals", "Medical Devices"]
assign_to: "Healthcare Sales"
- name: "Finance Specialist"
condition:
industry: ["Banking", "Insurance", "FinTech"]
assign_to: "Financial Services Sales"2. Round-Robin Distribution#
round_robin_config:
team: "SMB Sales"
members:
- name: Alice
capacity: 100%
max_leads_per_day: 20
- name: Bob
capacity: 100%
max_leads_per_day: 20
- name: Carol
capacity: 50% # Part-time
max_leads_per_day: 10
rules:
distribution: weighted # or equal
skip_if:
- out_of_office: true
- at_capacity: true
reset: daily
tracking:
log_assignments: true
balance_check: hourlyDistribution Algorithm:
┌─────────────────────────────────────────────────────────────┐
│ ROUND-ROBIN LOGIC │
├─────────────────────────────────────────────────────────────┤
│ │
│ 1. New lead arrives │
│ │ │
│ ▼ │
│ 2. Check team availability │
│ - Filter out: OOO, at capacity, off-hours │
│ │ │
│ ▼ │
│ 3. Calculate weighted position │
│ - Current assignments today │
│ - Capacity percentage │
│ - Last assignment time │
│ │ │
│ ▼ │
│ 4. Assign to rep with lowest weighted score │
│ │ │
│ ▼ │
│ 5. Update tracking, notify rep │
│ │
└─────────────────────────────────────────────────────────────┘3. AI-Powered Lead Scoring#
ai_scoring:
provider: openai
model: gpt-4
input_factors:
demographic:
- company_size
- industry
- job_title
- location
firmographic:
- annual_revenue
- employee_count
- funding_stage
- tech_stack
behavioral:
- pages_visited
- content_downloads
- email_engagement
- demo_requests
fit_score:
- icp_match_percentage
- competitor_usage
- budget_authority
scoring_prompt: |
Score this lead from 0-100 based on:
Our ICP (Ideal Customer Profile):
- B2B SaaS companies
- 50-500 employees
- Series A or later
- Using {competitor} or {similar_tool}
Lead Data:
{lead_data}
Return JSON:
{
"score": 0-100,
"fit_score": 0-100,
"intent_score": 0-100,
"tier": "A/B/C/D",
"reasoning": "...",
"recommended_action": "...",
"routing_suggestion": "..."
}
tier_thresholds:
A: 80-100 # Hot lead, immediate follow-up
B: 60-79 # Qualified, standard follow-up
C: 40-59 # Nurture, marketing sequence
D: 0-39 # Low priority, long-term nurture4. Territory Mapping#
territory_map:
north_america:
west:
states: [CA, WA, OR, NV, AZ, CO, UT]
owner: "West Coast Team"
reps: [Alice, Bob]
central:
states: [TX, IL, OH, MI, MN, WI]
owner: "Central Team"
reps: [Carol, David]
east:
states: [NY, MA, PA, FL, GA, NC]
owner: "East Coast Team"
reps: [Eve, Frank]
international:
emea:
countries: [UK, DE, FR, NL, ES, IT]
owner: "EMEA Team"
timezone: "Europe/London"
apac:
countries: [JP, SG, AU, KR, IN]
owner: "APAC Team"
timezone: "Asia/Tokyo"
overlap_resolution:
# When lead matches multiple territories
priority_order:
1: named_account_owner # If account already has owner
2: industry_specialist # If industry requires specialist
3: geography # Default to geography5. Workload Balancing#
workload_balancer:
check_frequency: hourly
metrics_tracked:
- current_open_leads
- leads_assigned_today
- leads_assigned_this_week
- average_response_time
- conversion_rate
balance_rules:
max_variance: 20% # Max difference between reps
rebalance_trigger:
- variance > max_variance
- rep_at_capacity
- rep_underperforming
rebalance_actions:
- pause_assignments: for_overloaded_rep
- increase_weight: for_underloaded_rep
- notify_manager: when_rebalancing
capacity_management:
per_rep:
max_open_leads: 50
max_new_per_day: 15
max_new_per_week: 60
team_level:
overflow_queue: true
overflow_notify: sales_manager
escalation_threshold: 2_hoursWorkflow Implementation#
Complete Lead Routing Workflow#
workflow: "Intelligent Lead Router"
trigger:
- type: hubspot_contact_created
- type: form_submission
- type: api_webhook
steps:
1. enrich_lead:
providers: [clearbit, zoominfo]
fields:
- company_size
- industry
- revenue
- location
- linkedin_url
2. score_lead:
method: ai_scoring
store_result:
hubspot_property: lead_score
3. determine_tier:
A_tier: score >= 80
B_tier: score >= 60
C_tier: score >= 40
D_tier: score < 40
4. apply_routing_rules:
sequence:
- check: named_account_owner
- check: industry_specialist
- check: territory_match
- check: round_robin_availability
5. assign_owner:
hubspot:
update_contact:
hubspot_owner_id: "{selected_owner_id}"
lead_status: "New"
lead_tier: "{tier}"
routing_reason: "{routing_logic}"
6. create_task:
hubspot:
type: CALL
subject: "Follow up: New {tier} lead - {company}"
due_date: "{sla_deadline}"
priority: "{priority_based_on_tier}"
notes: |
Lead Score: {score}
Routing Reason: {routing_reason}
Key Info: {summary}
7. notify_owner:
slack_dm:
message: |
🎯 *New Lead Assigned*
**{contact_name}** at **{company}**
Score: {score} ({tier} Tier)
📞 SLA: Respond within {sla_time}
Quick actions:
• [View in HubSpot]({hubspot_link})
• [LinkedIn]({linkedin_url})
• [Schedule Call]({calendly_link})
8. start_sla_timer:
deadline: "{sla_deadline}"
escalation_path:
- 50%_elapsed: reminder_to_owner
- 80%_elapsed: notify_manager
- 100%_elapsed: reassign + alertSLA Management#
sla_tiers:
tier_a:
response_time: 1_hour
escalation_path:
- 30min: slack_reminder
- 45min: manager_alert
- 60min: auto_reassign
tier_b:
response_time: 4_hours
escalation_path:
- 2h: slack_reminder
- 3h: manager_alert
- 4h: auto_reassign
tier_c:
response_time: 24_hours
escalation_path:
- 12h: slack_reminder
- 20h: manager_alert
- 24h: move_to_queue
sla_reporting:
metrics:
- response_time_avg
- response_time_p90
- sla_compliance_rate
- escalation_count
report_frequency: weekly
recipients: [sales_manager, ops_manager]Reporting Dashboard#
# Lead Routing Report - {Week}
## Distribution Summary
| Rep | Assigned | Responded | Avg Response | SLA Met |
|-----|----------|-----------|--------------|---------|
| Alice | 45 | 43 | 1.2h | 96% |
| Bob | 42 | 40 | 1.8h | 90% |
| Carol | 38 | 38 | 0.8h | 100% |
| **Total** | **125** | **121** | **1.3h** | **95%** |
## By Tier
| Tier | Count | Avg Score | Converted | Conv Rate |
|------|-------|-----------|-----------|-----------|
| A | 25 | 87 | 12 | 48% |
| B | 45 | 68 | 15 | 33% |
| C | 35 | 52 | 5 | 14% |
| D | 20 | 28 | 1 | 5% |
## Routing Breakdown
- By Territory: 60%
- By Industry: 25%
- Round Robin: 15%
## Issues
- 3 leads waited >SLA (reassigned)
- Alice at 95% capacity (monitor)
- No coverage for Healthcare vertical (gap)
## Recommendations
1. Hire Healthcare specialist
2. Increase Bob's training (response time)
3. Adjust A-tier threshold to 85 (too many false positives)Output Example#
Request: "Route this lead: John Smith, CTO at TechCorp (500 employees, SF, SaaS)"
Output:
# Lead Routing Decision
## Lead Profile
- **Name**: John Smith
- **Title**: CTO
- **Company**: TechCorp
- **Size**: 500 employees
- **Location**: San Francisco, CA
- **Industry**: SaaS
## AI Scoring
```json
{
"score": 85,
"fit_score": 90,
"intent_score": 80,
"tier": "A",
"reasoning": "Strong ICP fit - CTO at 500-person SaaS company in our target market. High authority buyer.",
"recommended_action": "Immediate outreach - high-value prospect"
}Routing Decision#
Assigned to: Alice Chen (Enterprise West)
Routing Logic:
- ✅ Territory: San Francisco → West Coast
- ✅ Company Size: 500 → Enterprise tier
- ✅ Industry: SaaS → No specialist needed
- ✅ Availability: Alice has capacity (18/20 today)
Action Items Created#
-
Task: Follow up call
- Due: 1 hour (Tier A SLA)
- Priority: High
-
Slack Notification: Sent to Alice
-
SLA Timer: Started (1h countdown)
Recommended Outreach#
Subject: Quick question about {pain_point} at TechCorp
Hi John,
Noticed TechCorp is scaling fast - congrats on the growth.
CTOs at similar SaaS companies often tell us {common_challenge}.
Would a 15-min call this week make sense to see if we can help?
[Calendly Link]
---
*Lead Routing Skill - Part of Claude Office Skills*FAQ
07 QUESTIONS
01What is the Lead routing agent skill?
Intelligent lead assignment and routing - AI-powered scoring, territory mapping, round-robin distribution, and workload balancing. It is listed under go-to-market in the 100xLab agent skills directory.
02How do I install Lead routing in Cursor, Claude Code, or Codex?
Run npx skills add claude-office-skills/skills --skill lead-routing in the project where you want it, then ask your agent for the skill by name. The --skill flag installs only Lead routing, not every skill in the repository.
03What does the Lead routing SKILL.md cover?
The SKILL.md includes sections on Overview, Routing Strategies, Rule-Based Routing, Round-Robin Distribution, AI-Powered Lead Scoring and Territory Mapping. Read the full text on this page or in plain markdown at /resources/agent-skills/claude-office-skills-skills--lead-routing--2qopyv/raw.
04Which agents can use Lead routing?
Lead routing installs with the cross-agent skills installer (npx skills add), which sets it up for agents such as Claude Code, Cursor and Codex. Each agent loads the SKILL.md when you ask for the skill by name.
05What license is Lead routing under?
The claude-office-skills/skills repository declares the MIT license. Read the license file in the repository before you copy or redistribute the skill.
06Is Lead routing free to use?
Reading, copying and installing Lead routing costs nothing: it is published in a public GitHub repository under the MIT license. Reuse and redistribution follow that license.
07Where is the source of Lead routing?
Lead routing comes from https://github.com/claude-office-skills/skills, which has 414 GitHub stars. The SKILL.md is at https://github.com/claude-office-skills/skills/blob/HEAD/lead-routing/SKILL.md. The author is claude-office-skills.
Need an agent workflow built around your stack?
Tell us the workflow you want automated. We reply within 4 working hours (IST).
ATTRIBUTION
Lead routing was written by its original authors and is published in claude-office-skills/skills (opens GitHub in a new tab) under the MIT license. All rights stay with the original authors.
100xLab did not write, maintain or endorse this skill. This page lists it for discovery. Repo stars belong to the source repository, not to the skill. Review any skill before you install it.