ClawPipe is a self-hosted AI agent pipeline platform. It provides CI/CD for AI workflows with task graphs, verifier layers, and durable memory. Production-ready, not a chat-UI wrapper.

How is ClawPipe different from LangGraph or CrewAI?

ClawPipe is fully self-hosted with built-in CI/CD primitives, durable state, and a verifier layer that catches hallucinations before they reach production. LangGraph and CrewAI are libraries; ClawPipe is a complete platform with dashboard, monitoring, and 36+ ready templates.

Can I run ClawPipe on my own infrastructure?

Yes. ClawPipe is self-hosted by design. Runs on Linux servers, Docker, or Kubernetes. No data leaves your infrastructure. EU and GDPR-compliant out of the box.

What does ClawPipe cost?

ClawPipe starts at €49/month for the Self-Hosted plan. Pricing scales by pipeline count and worker capacity. Enterprise pricing available for custom deployments.

Which AI models does ClawPipe support?

ClawPipe is model-agnostic. Works with Anthropic Claude, OpenAI GPT, Google Gemini, xAI Grok, Mistral, and local models via Ollama. Configurable per pipeline node.

Is ClawPipe production-ready?

Yes. ClawPipe has been running in production since 2025 across real estate, sales, and content automation workloads. Includes verifier layers, durable state, retry policies, and observability primitives required for production AI.

Private beta · v0.4

Pipelines for resilient AI agents.

36+ production pipelines you can deploy this week. Real estate, sales, content, devops, finance, compliance. Task graph + verifier + durable memory built in.

Browse pipelines → See pricing

clawpipe ▸ run property-watcher

// Define once. Re-run forever. const pipe = clawpipe.define({ id: "property-watcher", steps: [ { id: "fetch", fn: fetchListing, retries: 3 }, { id: "redact", fn: redact, deps: ["fetch"] }, { id: "verify", fn: verifySchema, deps: ["redact"] }, { id: "notify", fn: sendOwnerWA, deps: ["verify"], idempotent: true }, ], }); await pipe.run({ ref: "68899" });

● fetch · 88ms ● redact · 4ms (2 fields) ● verify · 12ms ● notify · 143ms

// What ClawPipe is

Six primitives that turn a chatbot into an agent.

Open-ended reasoning + deterministic workflows + verifier layer + durable memory + idempotent effects + trust-graded autonomy. Each one a separate concern, all wired together.

Task graph, not transcripts

Goals decompose into typed tasks with dependencies. Retries, fan-out, sub-tasks, approvals — first-class primitives, not prompt tricks.

Verifier layer

Planner → executor → verifier → optional human approval. Nothing ships until checks prove it shipped.

Durable memory

BM25 + vector + RRF fusion across artifacts, decisions, runs. Inspectable. Editable. Versioned. Dedup-aware.

Idempotent effects

HTTP, WhatsApp, email, payments. Idempotency keys, retry policies, compensating actions. No double-sends.

Durable waits

Pause for minutes, hours, or days without losing run state. Resume exactly where it left off — no manual stitching.

Trust-graded autonomy

Skills earn autonomy through measured production success. Approvals stay where blast radius is real.

// How it runs

Every run, the same five rails.

Goal in. Artifact out. Every step traced, scored, and re-runnable from any phase.

Goal

Plan

Execute

Verify

Ship

      checkpoints · resume from any phase
      artifacts · every step writes a file
      trace · full trajectory, not just outcome
      eval · pass-rate tracked per skill
    

// Stack

Adapters over lock-in.

Bring your own model, vector store, queue, and side effects. ClawPipe normalizes the contract — you keep ownership.

Models

Routed by complexity, latency, and budget. Fall back to local when external is down.

Claude
GPT-5
Gemini
Ollama
vLLM

Memory

SQLite + FTS5 + vector embeddings out of the box. Swap to pgvector, Qdrant, or LanceDB at scale.

SQLite FTS5
MiniLM-L6 384d
pgvector
Qdrant

Effects

HTTP, WhatsApp, email, calendar, Stripe, GitHub. Idempotency keys + retries + compensations baked in.

WhatsApp
SMTP
Stripe
GitHub
HTTP

Runtime

Self-hosted Node.js + SQLite. Add Postgres + Redis when you need horizontal scale. No vendor lock-in.

Node 22
SQLite
Postgres
Redis
PM2

// Pricing

Pick the plan that ships work.

Start with a single pipeline. Scale to a full ops stack. Cancel anytime.

Starter

€49/mo

For solo operators running 1-2 pipelines.

✓ 3 active pipelines
✓ 10k runs/month
✓ Email support
✓ All registry templates

Start →

What ClawPipe ships in production.

Real loops running real ops. Names anonymized — outcomes verified.

Real estate · Mediterranean

11× faster buyer matching

Agency was manually matching buyer briefs against 1,500 live listings. Avg time per buyer: 35 min, miss rate ~80%.

ClawPipe wired: form → semantic re-rank → hard filters (date, location, budget, lifestyle) → per-lead share link. Avg now: 3 min, hit rate >70%.

🚀 11× faster

🎯 +340% hit rate

⏱ 3 weeks to live

Compliance · AML SaaS

KYC onboarding from 3 days to 3 hours

SaaS was onboarding clients manually. Sanctions screen, PEP check, risk scoring done in 6 separate tools.

ClawPipe pipeline: OCR → screen → score → human approval gate for amber/red → audit trail. Same SLA, automated.

📉 -95% time

🛡 100% audit trail

👥 3 ops freed

Pipelines for resilient AI agents.

Modern pipelines for AI-native ops.

Six primitives that turn a chatbot into an agent.

Task graph, not transcripts

Verifier layer

Durable memory

Idempotent effects

Durable waits

Trust-graded autonomy

Every run, the same five rails.

Adapters over lock-in.

Models

Memory

Effects

Runtime

Pick the plan that ships work.

What ClawPipe ships in production.

11× faster buyer matching

KYC onboarding from 3 days to 3 hours

Tell us what you want to ship.

See what's possible.

Guides for AI engineers

Agent Pipelines vs LangGraph

AI Workflow CI/CD Explained

Self-Hosted Agent Platforms 2026

Task Graph vs Chat Agent

Production-Ready AI Checklist

All guides →