maleekduckett

Deal Risk Agent — Data skill for Claude Code

Data community

multi-agent deal risk orchestration scoring using n8n.

How to install Deal Risk Agent

This entry records only its repository, not the path inside it, so there is no exact command to give. Open maleekduckett/deal-risk-agent and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Deal Risk Agent does

multi-agent deal risk orchestration scoring using n8n. 3 Claude subagents, narrow in scope and highly specialized along with an additional synthesis agent that as the name implies, synthesizes data from the 3 subagents. 100% synthetic pipeline data, claude assisted

Alternatives in Data

n8n Code JavaScript — JavaScript in n8n Code nodes with data access patterns 3.6k ★
Portaljs Add Resource — Add another file (resource) to an EXISTING dataset in a PortalJS portal — a data dictionary, methodology, or a 2.3k ★
Claude Deep Research Skill — Enterprise-grade deep research skill for Claude Code with 8-phase pipeline, source credibility scoring, and au 1k ★

README

Deal Risk Agent

A multi-agent workflow that reviews open deals and flags which ones are at risk, why, and what the rep should do next. Built in n8n with Claude, run on 50 synthetic deals.

The problem

Pipeline reviews usually come down to a manager reading deal notes and going with their gut. Risk signals are spread across activity data, deal fit, and notes about competitors, and nobody has time to check every deal the same way every week.

This workflow checks each deal from three angles, then combines them into one risk call with a recommended next step.

How it works

![The deal risk workflow in n8n](docs/n8n-workflow.png)

flowchart LR
    A[Deals sheet] --> B[Run Context
unique run ID]
    B --> C[Read Deals]
    C --> D[Engagement Agent]
    C --> E[ICP Fit Agent]
    C --> F[Competitive Agent]
    D --> G[Merge on deal_id]
    E --> G
    F --> G
    G --> H[Synthesis Agent]
    D -.-> L[(Run log sheet)]
    E -.-> L
    F -.-> L
    H -.-> L

Each agent runs on Claude Sonnet 5 at low temperature (0.1) and returns structured JSON, so every output has the same shape and can be logged and analyzed.

The agents

Each agent is a specialist with one job, a fixed set of inputs, and a fixed output shape.

Agent	Specialty	Sees only	Returns
Engagement	Buyer engagement analyst. Judges whether the buyer is active or going quiet.	Company name, email count, meeting count, average response lag, deal notes	Risk (Low / Medium / High), confidence, one-sentence reason
ICP Fit	Ideal customer profile analyst. Judges whether this is the right kind of deal.	Company name, industry, deal amount, contact title, plus the ICP rules ($50K+ deal, Director-level or above buyer, no startups, government, or education)	Fit (Strong / Moderate / Weak), confidence, one-sentence reason
Competitive	Competitive intelligence analyst. Looks for competitor names, price comparisons, and competing bids.	Co