Everything you need to
ship software with AI agents

The Delivery Engine plans the work, your agents build in the cloud, and your team stays in control. Here is what that looks like, tier by tier.

Free tier

The core platform

The Delivery Engine

Your agents plan before they build.

Every task becomes a phased plan you review before anything runs. Your agents then work through the phases in parallel, so features that used to take days land in hours.

You approve the plan before a single line is written.
Independent phases run in parallel. Multiple tasks at once.
Every phase has clear objectives and acceptance criteria.
Opinionated presets out of the box. You pick the implementation, not the scaffolding.

Cloud execution

Build from anywhere, on any device.

Your agents run in isolated cloud sandboxes. Multiple tasks at once, from your laptop, tablet, or phone. No local setup, no GPU, no rate limits on your machine.

Isolated sandboxes per run. Your code never shares state with another tenant.
Multiple plans execute in parallel across your project.
Works from any device with a browser. Close your laptop, your agents keep building.
Mobile app with Siri and voice integration lets you kick off work on the go.

Cloud Execution 3 sandboxes

Rate limiting Building

Auth tokens Building

Dashboard v2 Complete

Laptop

Tablet

Phone

Code review

Review like any PR.

A built-in file browser and diff viewer let you review every change your agents wrote. Auto-PR generation means your work lands as a real pull request, ready for your existing workflow.

File browser with syntax highlighting and full directory navigation.
Inline diff viewer showing exactly what changed and why.
Auto-generated pull requests merge into your existing Git workflow.
Review comments feed back into the next plan, closing the loop.

Files

rateLimit.ts

tokenValidator.ts

rateLimit.test.ts

æ feat: rate limiting

#142 Open

14 export class RateLimiter {

15 + private bucket: TokenBucket;

16 + private window = 60_000;

17 - private limit = Infinity;

18 + private limit = 100;

19 }

✓ CI passing +276 -12

Project management

Project management, built in.

Work lives in an issues, domains, and features model with a real lifecycle: backlog, planning, in progress, review, complete. It works like the project tools you already know.

AI models

Models, managed for you.

Weaver picks the right model for each task and runs it for you. No provider setup, no API keys to manage. Want direct control? Bring your own API key at any time.

Basic tier

Agents that learn. Code that gets tested.

Knowledge base

Gets smarter every time you use it.

Every decision you make, every convention you set, every piece of context you add makes every future plan sharper. Your agents don't start from scratch. They start from everything your team has already learned.

Persist architecture decisions, conventions, and domain knowledge once.
Your agents pull the right context into each plan automatically.
Knowledge docs are co-editable, so the source of truth stays current.
Memory compounds. Every shipped feature makes the next plan sharper.

Knowledge Base 14 entries

𝔻 Auth conventions Architecture 3d ago

𝔻 Rate-limit policy API 1d ago

𝔻 Test naming rules QA 4h ago

Agent syncing

12 decisions captured

QA and testing

Tests that actually stick.

Coverage dashboards, flaky test detection, and test-gap scanning give your agents and your team visibility into test health. Your agents write tests as part of every plan, not as an afterthought.

Coverage dashboard shows exactly where your test suite is thin.
Flaky test detection flags unreliable tests before they erode trust.
Test-gap scanning identifies untested code paths across your project.
Every plan includes test criteria. Your agents write and verify tests as they build.

Test Suite 42/44 passing

Coverage 78%

✓ auth.login.test.ts 1.2s

⚠ rateLimit.test.ts Flaky 3.8s

✓ billing.charge.test.ts 0.9s

2 test gaps detected

Scholar

Multi-agent codebase intelligence.

Scholar deploys a team of specialized agents to mine your codebase. They map architecture, extract patterns, surface dependencies, and generate structured briefs that feed directly into your knowledge base.

Pro tier

Your whole team, in one workspace.

Collaborative documents

Write together, then hand it to your agents.

Real-time editing with live cursors, inline comments, and spec generation. Draft a spec collaboratively, then let your agents turn it into a phased plan and build it.

Live cursors and presence show who is editing what, in real time.
Inline comments turn document review into a conversation.
Spec generation transforms a document into a structured plan.
Documents link to plans, features, and knowledge base entries.

API Spec v2

AR JT

2 editing

Rate-limit middleware specification

Apply a token-bucket limiter to all public API Ana

routes. Sliding window of 60 seconds, 100 req Jake

default. Override per-route via config.

Should we make the window configurable per-tenant?

Production context

Agents that know your live system.

Your agents see more than source files. Weaver feeds them live system context: error tracking, logs, health, and dashboards. They reason about the running system, not a frozen snapshot.

Wire in error tracking, logs, metrics, and dashboards as first-class context.
Your agents cite real incidents and telemetry when they plan a fix.
Close the loop from a production signal to a reviewed, tested change.
Context your code editor never sees, because it lives beyond the repo.

Live-system context

streaming

Sentry TypeError · checkout.ts · 142 events
Logs p99 latency 1.8s on /api/plan
Health worker pool 3/4 · queue depth 12
Dashboards conversion down 4.2% since deploy

Agent context

Your agents plan against the running system, not a stale snapshot.

Voice and video

Talk through a plan, then build it.

Meeting rooms with live transcription and AI-generated summaries. Discuss a feature, review a plan, or pair on a problem. The transcript becomes context your agents can act on.

Team collaboration

Work together, in real time.

Presence, co-editing, comments, and shared chat make Weaver a place your whole team works together. Review a plan, leave a comment, and hand off. Everyone sees the same state.

Enterprise

Security, compliance, and scale.

Identity and access

SSO and SAML for your org.

Single sign-on and SAML integration for your identity provider. Enforce org-wide access policies and provision users from your directory.

Audit and compliance

Full audit trail, exportable.

Every action, approval, and change is logged. Export audit data for compliance reviews, or stream it to your SIEM. A signed DPA is available on request.

Support and billing

Dedicated support with an SLA.

Priority support, a guaranteed SLA, and net-30 invoicing that fits enterprise procurement. Volume model rates scale with your org.

Talk to sales

Ready to try it?

The Delivery Engine, project management, a knowledge base, QA, and team collaboration. All in one platform.

Get started free

Everything you need to ship software with AI agents