Twill

Twill is a software factory as a service for engineering teams. It ingests signals from across your stack and turns them into verified pull requests, using isolated task forks for testing and review before production.

Outils Dev IA Assistant DevOps IA Agent de programmation

Visiter le Site Web

Overview

Twill is presented as a software factory as a service for engineering teams. Its homepage describes a system that ingests signals from across a stack and turns them into verified pull requests that are checked before they reach production.

The product centers on automated software delivery workflows. Tasks can begin from GitHub, Slack, or Linear activity, then run in an isolated copy of the company environment where code, databases, tests, and previews are available for verification. The page also says Twill can operate on demand, on a schedule, or from event-driven triggers.

Core capabilities

Multi-repo context

Connect multiple repositories so an agent can reason across frontend, backend, workers, infrastructure, and shared packages in one task.

Full task environment

Run work inside an isolated fork where agents can install packages, start services, seed data, execute tests, and inspect previews.

Ready-to-run environment

Start from a cloned, prepared developer environment with dependencies installed and services already warm, reducing setup time before the task begins.

Flexible triggers

Launch tasks from GitHub, Slack, or Linear signals, including on-demand requests, scheduled automations, and event-based triggers.

Model routing with your keys

Use different models for different jobs, with the site calling out Claude, GPT, Qwen, Kimi, and GLM and stating that teams can use their own keys at provider rates.

Workflow integrations

Connect existing tools or extend workflows with MCP servers and skills, alongside the integrations named on the homepage.

Practical workflows

Incoming work triage
Use Twill to triage incoming GitHub, Slack, or Linear signals and turn them into a task that runs in a prepared environment instead of starting from scratch.
Recurring automation
Schedule recurring engineering chores such as incident follow-up, cloud-resource checks, documentation updates, and issue follow-up PRs.
Cross-repository changes
Work across multiple repositories when a change spans application code, workers, infrastructure, and shared packages, so the task has access to the full stack context.
Verified PR delivery
Let the system run tests and inspect previews in an isolated fork before a pull request is shipped, which supports verification-heavy workflows.
Custom workflow integration
Connect Twill to existing team tools or extend it with MCP servers and skills when a standard integration is not enough.

Pros and Cons

Pros

Supports automated work from multiple trigger types, including on-demand, scheduled, and event-based runs.
Can reason across multiple repositories, which is useful for larger stacks split across frontend, backend, infrastructure, and shared packages.
Runs tasks in an isolated environment with code, services, tests, and screenshots available for verification.
Supports multiple model options rather than locking teams to one provider.
Offers a free Pro subscription for open source projects, capped at the Pro plan with no overage.

Cons

The public site provides only high-level capability descriptions, so setup, limits, and workflow details remain unclear.
Pricing is not described on the homepage, beyond a free Pro subscription offer for open source projects.
The site lists integrations and trigger sources but does not document configuration depth or coverage for each tool.

FAQ

How does Twill start work?

Twill ingests signals from tools across your stack and can launch work on demand, on a schedule, or from triggers such as a GitHub PR, Slack message, or Linear issue.

What happens when a task runs?

The homepage says Twill can run in isolated task forks with the company environment already cloned, dependencies installed, services running, and tests and screenshots available for verification.

What kinds of teams is Twill for?

Twill is positioned for engineering teams that want to automate recurring software delivery work, including triage, infrastructure checks, documentation updates, and opening PRs when something needs to change.

Which tools does Twill connect to?

The site mentions GitHub, Slack, Linear, Notion, Sentry, Google Cloud, AWS, Asana, and Datadog, and also says you can add your own workflow with MCP servers and skills.

Is pricing or setup information available on the homepage?

The homepage presents Twill as a software factory as a service, but it does not provide detailed plan information, setup requirements, or documented limits on the public page.

Quick Facts

Category: Developer Tool
Product type: Software factory as a service
Primary users: Engineering teams
Trigger sources: GitHub, Slack, Linear, schedules, and automations
Integrations mentioned: Notion, Sentry, Google Cloud, AWS, Asana, Datadog, GitHub, Slack, Linear
Website: trysummon.com