Skip to content
312+ businesses automated avg. 14h/week savedManual workflows cost the average team £512/week fix it in 10 daysDeployed in 5–10 business days · 30-day money-back guaranteeDental · Real Estate · Agencies · E-commerce · Covered99.97% uptime SLA · Monitored 24/7 by our ops teamA full-time ops hire costs £45K+/yr PURIST delivers more in daysn8n · Make · Claude AI · 500+ workflow templatesFree automation audit limited to 5 spots this week312+ businesses automated avg. 14h/week savedManual workflows cost the average team £512/week fix it in 10 daysDeployed in 5–10 business days · 30-day money-back guaranteeDental · Real Estate · Agencies · E-commerce · Covered99.97% uptime SLA · Monitored 24/7 by our ops teamA full-time ops hire costs £45K+/yr PURIST delivers more in daysn8n · Make · Claude AI · 500+ workflow templatesFree automation audit limited to 5 spots this week312+ businesses automated avg. 14h/week savedManual workflows cost the average team £512/week fix it in 10 daysDeployed in 5–10 business days · 30-day money-back guaranteeDental · Real Estate · Agencies · E-commerce · Covered99.97% uptime SLA · Monitored 24/7 by our ops teamA full-time ops hire costs £45K+/yr PURIST delivers more in daysn8n · Make · Claude AI · 500+ workflow templatesFree automation audit limited to 5 spots this week
PURIST
312+
Clients automated
14 h/wk
Avg time saved
99.97%
Uptime SLA
< 7 days
Deploy time
PURIST AI
Claude Opus 4.7 · n8n v1.71 · <80ms
What type of business are you running? I'll show you exactly which processes we'd automate first and your estimated ROI.
Powered by n8n + Claude Opus 4.7 Book free audit →
Documentation Hub

Everything you need,
in one place.

Platform guides, integration references, AI documentation, and deployment best practices. From getting started to scaling production.

PURIST

Automation-as-a-Service platform

PURIST deploys, manages, and monitors n8n + Claude AI automation stacks for businesses. We handle the infrastructure so you focus on growth.

Getting started

What is PURIST?

PURIST is an automation-as-a-service company. We design, build, deploy, and maintain custom automation workflows for your business using n8n and Claude AI. You get a fully managed stack — no code to write, no servers to maintain, no tools to learn.

How the onboarding works

It starts with a free 45-minute automation audit where we map your manual processes. Within 5 business days, you receive a deployment plan. After approval, your first workflows go live within 10–15 days. You get a live dashboard from day one.

What industries we serve

Professional services, e-commerce, finance, healthcare admin, recruitment, marketing agencies, and SaaS companies. Any business with repetitive manual processes is a fit. We have deployed 312+ stacks across 14 industries.

Pricing structure

PURIST charges a one-time setup fee (based on complexity) plus a monthly retainer for monitoring, maintenance, and iteration. Setup fees typically range from £1,500–£5,000. Monthly retainers start at £300/month. No per-workflow charges, no usage caps.

Platform & infrastructure

Your stack, your servers

Every client gets a dedicated n8n instance running on isolated infrastructure. We do not commingle client data. Your workflows, credentials, and execution logs live on your own stack — we manage it, you own it.

Uptime & reliability

99.97% uptime across all client deployments, measured over a rolling 12-month window. Each stack includes automated health checks every 60 seconds, dead-man switches on critical workflows, and instant failover.

Security model

All credentials are encrypted at rest (AES-256) and in transit (TLS 1.3). We use scoped API tokens with rotation every 90 days. HMAC signature validation on all incoming webhooks. SOC 2 Type I compliance in progress.

Monitoring & alerting

Every workflow execution is logged with full input/output capture. Failures trigger instant Slack/email alerts. You get a real-time dashboard showing execution volume, success rates, latency, and error trends.

Workflows & integrations

What can be automated?

Lead capture and CRM entry, invoice generation, email sequences, report compilation, data synchronisation between tools, client onboarding flows, social media scheduling, document generation, and any process that follows a repeatable pattern.

Supported integrations

PURIST supports 500+ integrations via n8n nodes: Slack, HubSpot, Salesforce, Xero, QuickBooks, Google Workspace, Microsoft 365, Notion, Airtable, Stripe, Shopify, and any tool with an API or webhook.

Custom workflows

Every PURIST deployment is custom-built. We do not use templates. Your workflows are designed around your specific tools, team structure, and business logic. We handle edge cases, error recovery, and conditional branching.

AI-powered workflows

Claude AI is embedded in workflows for tasks that need judgement: classifying support tickets, drafting email replies, summarising meeting notes, extracting data from unstructured documents, and making routing decisions.

Support & SLA

Response times

Critical issues (workflow down): 1-hour response, 4-hour resolution target. Standard requests: same-business-day response. Feature requests and optimisations: included in your monthly retainer, delivered in weekly iteration cycles.

Dedicated Slack channel

Every client gets a shared Slack channel with the PURIST engineering team. No ticket queues, no support tiers. You message us directly and we respond.

30-day guarantee

If your automation stack does not deliver measurable time savings within 30 days of deployment, we refund your setup fee in full. No questions, no process.

Iteration & scaling

Your stack evolves with your business. Monthly retainer includes ongoing optimisation: new workflows, workflow improvements, integration additions, and scaling adjustments. No change request fees.

Claude Skills

329 open-source prompt skills for Claude

Claude Skills is PURIST's open-source library of 329 structured prompt skills that extend Claude's capabilities for professional use cases across engineering, marketing, product, operations, and compliance.

Overview

What are Claude Skills?

Claude Skills are structured prompt templates that give Claude domain-specific expertise. Each skill defines a role, methodology, output format, and quality standard. They transform Claude from a general assistant into a specialist — copywriter, code reviewer, financial modeller, or compliance auditor.

How skills work

Each skill is a markdown file that you add to your Claude configuration. When activated with the @ prefix (e.g. @CopywriterAgent), Claude adopts the full persona: methods, frameworks, output structure, and quality bar. Skills persist across the conversation until changed.

Skill categories

5 domains, 20 teams, 100 agent-level skills covering: Marketing & Growth (20 skills), Business & Strategy (20), Operations & Productivity (20), Tech & Product (20), and Intelligence & Research (20). Plus 229 utility skills for specific tasks.

Open source licence

All 329 skills are MIT-licensed and free to use, modify, and distribute. No attribution required for commercial use. The skills library is maintained by PURIST and accepts community contributions.

Installation & setup

Claude Code setup

Add skills to your Claude Code configuration by placing the skill markdown files in your .claude/skills/ directory or referencing them in your CLAUDE.md file. Skills are loaded automatically at conversation start.

Claude.ai (web) setup

Copy the skill prompt into a Claude Project's custom instructions. Each Project can hold multiple skills. For the full 100-agent library, add the complete prompt to a single Project.

API usage

Include skill definitions in the system prompt when calling the Claude API. Skills work with all Claude models (Opus, Sonnet, Haiku). For production use, we recommend Sonnet for speed-sensitive workflows and Opus for complex reasoning tasks.

Customising skills

Skills are plain markdown — edit them to match your brand voice, industry terminology, or specific frameworks. Fork a skill, modify the methodology section, and it becomes your own proprietary variant.

Skill catalogue highlights

Marketing & Growth

TrendScout (trend detection), CopywriterAgent (direct-response copy), SEOAgent (technical SEO), GrowthHacker (AARRR funnel), EmailMarketer (lifecycle sequences), AdOptimizer (Meta/Google/LinkedIn ads), VideoScriptAgent (YouTube/TikTok scripts), LaunchAgent (go-to-market).

Business & Strategy

MarketAnalyst (TAM/SAM/SOM), CompetitorIntel (battlecards), FinancialModeler (unit economics), PricingAgent (tier strategy), InvestorDeck (pitch narratives), NegotiationAgent (Harvard + Chris Voss methods), OKRAgent (outcome-driven KRs).

Tech & Product

ProductManager (PRDs + RICE scoring), CodeReviewer (principal-level reviews), ArchitectAgent (system design + ADRs), SecurityAgent (STRIDE threat modelling), PerformanceAgent (bottleneck analysis), DevOpsAgent (CI/CD pipelines), TestingAgent (test pyramids).

Operations & Intelligence

ProjectManager (Agile/Waterfall), MeetingSummarizer (transcription → actions), DataAnalyst (cohort analysis), ProcessOptimizer (Lean Six Sigma), ResearchAgent (systematic reviews), PredictiveAgent (forecasting), HypothesisAgent (MECE issue trees).

Best practices

One skill per task

Activate one skill at a time for best results. Each skill sets a specific persona and methodology. Mixing skills in one message reduces output quality. Switch skills with a new @ prefix.

Provide context

Skills work best when you give them the context they need: your industry, target audience, constraints, and goals. The more specific your input, the more targeted the output.

Iterate, don't restart

Skills maintain context across the conversation. Ask for revisions, alternatives, or deeper dives on specific sections rather than starting over. The skill builds on everything discussed.

Chain skills for complex projects

For large projects, use ChiefOrchestrator to plan which skills to deploy in sequence. Example: ResearchAgent → ProductStrategist → UserStoryAgent → SprintAgent for a full product cycle.

Claude AI

The AI engine behind PURIST workflows

Claude is Anthropic's frontier AI model, embedded in every PURIST deployment for tasks that require reasoning, language understanding, and decision-making. Here's what you need to know.

Core capabilities

Language understanding

Claude processes and generates text at near-human quality. In PURIST workflows, this means: reading and classifying inbound emails, drafting personalised responses, summarising documents, and extracting structured data from unstructured text.

Tool use (function calling)

Claude can call external tools and APIs during a conversation or workflow. This powers automated actions: looking up a CRM record, creating a calendar event, updating a spreadsheet, or triggering a downstream workflow — all decided by Claude in real time.

Vision & multimodal

Claude can read images, screenshots, charts, and documents. In PURIST workflows, this enables: processing scanned invoices, reading receipts, analysing dashboard screenshots, and extracting data from PDFs with mixed text and images.

Extended thinking

For complex decisions, Claude uses extended thinking — a chain-of-thought process that reasons through multi-step problems before responding. This powers: financial analysis, compliance reviews, and strategic recommendations.

Models & performance

Claude Opus 4

The most capable model. Used in PURIST workflows for complex reasoning: contract review, financial modelling, strategic analysis, and multi-step decision chains. Higher cost per token, but dramatically better on hard tasks.

Claude Sonnet 4

The workhorse model. Best balance of speed, cost, and intelligence. Used in 80% of PURIST workflows: email classification, data extraction, response drafting, and standard decision routing. Sub-2-second response times.

Claude Haiku

The fastest model. Used for high-volume, low-complexity tasks: quick classification, simple extraction, and routing decisions where speed matters more than depth. 10x cheaper than Opus.

Model selection in PURIST

We choose the right model for each node in your workflow. A single workflow might use Haiku for initial classification, Sonnet for drafting, and Opus for final quality review. You pay for intelligence where it matters.

Using Claude in automation

Prompt engineering

PURIST writes production-grade prompts for every Claude node. We use structured output schemas, few-shot examples, and system prompts tuned for consistency. Every prompt is version-controlled and A/B tested against real workflow data.

Structured outputs

Claude returns JSON, XML, or structured text that downstream nodes can parse reliably. We define output schemas per task — no free-form text that breaks your pipeline. Schema validation catches format errors before they propagate.

Error handling & fallbacks

If Claude returns an unexpected result, the workflow retries with a clarified prompt. After 3 attempts, the task is flagged for human review. No silent failures, no broken data.

Cost management

We optimise token usage across your workflows: concise prompts, cached system messages, and model-appropriate task assignment. Average Claude cost per client: £40–120/month depending on volume.

Anthropic

The company behind Claude

Anthropic is the AI safety company that builds Claude. Understanding their approach helps you understand why PURIST chose Claude as our AI backbone.

Company & mission

About Anthropic

Founded in 2021 by former OpenAI researchers, Anthropic is an AI safety company building reliable, interpretable, and steerable AI systems. They are backed by Google, Spark Capital, and others, with over $7B in funding.

Why PURIST uses Anthropic

Three reasons: (1) Claude is the most reliable model for structured business tasks — lowest hallucination rate in independent benchmarks. (2) Anthropic's safety-first approach means predictable, controllable outputs. (3) Enterprise-grade API with 99.9% uptime SLA.

Constitutional AI

Anthropic's training approach, Constitutional AI, teaches Claude to be helpful, harmless, and honest through a set of principles rather than human feedback alone. This makes Claude more consistent and less likely to produce harmful or unreliable outputs.

Research & transparency

Anthropic publishes peer-reviewed research on model interpretability, scaling laws, and safety techniques. Their published papers inform how PURIST configures and deploys Claude for maximum reliability.

API & platform

API access

The Claude API is available at api.anthropic.com. PURIST manages all API access on your behalf — you never need to create an Anthropic account or manage API keys. We handle authentication, rate limiting, and billing.

Rate limits & scaling

Anthropic provides tiered rate limits based on usage. PURIST maintains enterprise-tier access ensuring your workflows never hit rate limits. For high-volume deployments, we implement request queuing and load balancing.

Data privacy

Anthropic does not train on API inputs. Your business data sent through PURIST workflows is not used to improve Claude models. Anthropic retains API logs for 30 days for abuse monitoring, then deletes them.

Console & monitoring

The Anthropic Console (console.anthropic.com) provides usage dashboards, API key management, and billing. PURIST clients receive a monthly usage report with cost breakdowns per workflow.

Resources

Official documentation

Anthropic's docs at docs.anthropic.com cover the full API reference, prompt engineering guides, tool use documentation, and model comparison tables. PURIST's deployment practices are built on these official guidelines.

System status

Check status.anthropic.com for real-time API availability. PURIST monitors this feed and proactively reroutes traffic during degradations. We maintain fallback configurations for all critical workflows.

Model updates

Anthropic releases model updates regularly. PURIST tests all new model versions against your workflows before upgrading. We never auto-upgrade production deployments — every model change is validated and communicated.

Enterprise support

Anthropic offers enterprise plans with custom rate limits, dedicated support, and SLAs. PURIST manages the enterprise relationship on behalf of high-volume clients, simplifying procurement.

n8n

The workflow engine powering PURIST

n8n is the open-source workflow automation platform at the core of every PURIST deployment. It connects your tools, executes your logic, and handles the orchestration layer.

What is n8n?

Overview

n8n is a source-available workflow automation platform with 500+ integrations, a visual workflow editor, and a powerful execution engine. Unlike Zapier or Make, n8n can be self-hosted — meaning your data stays on your infrastructure.

Why PURIST chose n8n

Self-hosting means client data never leaves their infrastructure. 500+ native integrations cover 95% of use cases without custom code. The visual editor makes workflows auditable by non-technical stakeholders. And it scales to thousands of executions per hour.

n8n vs Zapier

n8n: self-hosted, no per-task pricing, full data control, complex branching, code nodes. Zapier: cloud-only, per-task billing (adds up fast), simpler but limited. For serious automation, n8n wins on cost, control, and capability.

n8n vs Make

n8n: open source, self-hosted, native code execution, better error handling. Make: proprietary, cloud-only, visual-first but limited debugging. PURIST uses n8n for its transparency and infrastructure control.

Core concepts

Workflows

A workflow is a sequence of connected nodes that execute in order. Each node performs one action: fetch data, transform it, send it somewhere, or make a decision. Workflows can branch, loop, and handle errors at every step.

Nodes

Nodes are the building blocks. There are trigger nodes (start the workflow), action nodes (do something), and function nodes (run custom code). PURIST uses ~40 node types across typical deployments.

Credentials

n8n stores API keys and OAuth tokens encrypted. PURIST manages credential rotation, scoping, and access control. Each workflow only has access to the credentials it needs — principle of least privilege.

Executions

Every time a workflow runs, it creates an execution record with full input/output data for every node. PURIST retains execution data for 30 days and provides searchable logs through your monitoring dashboard.

Integrations

CRM & Sales

HubSpot, Salesforce, Pipedrive, Close, Copper, Zoho CRM. Bi-directional sync, lead routing, deal updates, contact enrichment, and activity logging — all automated through native n8n nodes.

Finance & Accounting

Xero, QuickBooks, Stripe, PayPal, GoCardless, Wise. Invoice generation, payment reconciliation, expense categorisation, and financial reporting workflows.

Communication

Slack, Microsoft Teams, Gmail, Outlook, Twilio, SendGrid. Message routing, notification workflows, email sequences, and multi-channel alert systems.

Productivity & Data

Google Sheets, Airtable, Notion, Monday.com, Asana, Jira, Trello, PostgreSQL, MySQL, MongoDB. Data sync, task creation, project tracking, and database operations.

Self-hosting & infrastructure

Deployment model

PURIST deploys n8n on dedicated infrastructure per client. Options include AWS, GCP, Azure, or your own servers. Each instance is containerised (Docker), version-controlled, and backed up daily.

Scaling

n8n supports worker mode for high-volume deployments: separate the web UI from execution workers. PURIST configures auto-scaling based on execution queue depth. Most clients run comfortably on a single instance.

Updates & maintenance

PURIST handles all n8n version upgrades. We test new releases in a staging environment against your workflows before production deployment. Zero-downtime updates using blue-green deployment strategy.

Backups & disaster recovery

Daily automated backups of all workflows, credentials (encrypted), and execution data. 30-day retention. Point-in-time recovery available within 4 hours. Backup verification tests run weekly.

Ready to automate?

Book a free 45-minute automation audit. We'll map your manual processes and show you exactly what can be automated — and what it will save you.