# YAGNI > Coding agents that know your company. YAGNI is a drop-in CLI and desktop coding agent built for engineering teams: a shared decision ledger, agents grounded in your business tools with citations, per-developer budgets, and intelligent routing across vetted, US-hosted open weight models at 60% or more under comparable frontier API rates. Behind the harness, agent Teams carry whole engineering scopes and earn autonomy on the record; nothing consequential ships without a person's approval. Additive to your stack. The door is a 30-minute call with the founder. YAGNI ("You Ain't Gonna Need It") is the coding agent your engineers already know how to run: same commands, same muscle memory, and your skills, plugins, and MCPs slot straight in. One /go run moves through named stages, map, plan, build, adversarial review, pull request, and pauses at the checkpoints that are yours: nothing is built until you approve the plan, the reviewer attacks the change through up to three fix rounds before the PR opens, and the meter prints the cost of every step as it happens. What makes it a team harness rather than a personal tool: every run writes to a shared decision ledger (what was chosen, what was rejected, and why), so agents stop re-deciding what your team already settled and engineers read the record instead of asking around. When one developer corrects an agent, every developer's agents learn it. Usage and budgets are visible per developer. Thirty days in, YAGNI knows how your team builds software, and everything it learned is yours. The same machinery, held by an agent Team with Responsibilities and a record, carries whole lanes of engineering work: PR review through the lenses your team sets, production watch, incident triage, answering "why is this built that way," turning threads into well-formed tickets, and proposals argued against the one Number the Team owns. Consequential work stages a Decision and waits for the person authorized to answer; every action leaves a Receipt proved from the source. Autonomy climbs a ladder (Training, Supervised, Autonomous) earned on the track record and confirmed by a person, and irreversible actions stay behind a person's approval at every level, forever. YAGNI is additive. It reads the tools you already pay for and ships approved work back into them; it never hosts your apps or migrates your data. There is no self-serve funnel: the workspace is set up on a 30-minute onboarding call (GitHub App installed with your security lead present, the first Teams built by hand in the room), and teams pilot YAGNI Code free with two or three developers, reading the cost comparison from their own sessions. ## Product - [How the engineering org works](https://yagni.app/code): The mechanics: the loop that verifies outcomes instead of stopping at the PR, what the org runs day one, and the ladder from review to approval. - [Pricing](https://yagni.app/pricing): YAGNI Code is metered, not per seat: the whole team is in the workspace at no charge and the floor is 60% or more under comparable frontier API rates. The org product is sold in Teams and never metered. No public rate card; the number is scoped on the call. - [Book 30 minutes](https://yagni.app/book): The one door. Thirty minutes with the founder; bring a usage export from your current coding agent. - [Security](https://yagni.app/security): Workspace-scoped data access, encrypted connector credentials, sanitized action auditing, and US-hosted inference under zero data retention agreements. SOC 2 Type II certification is in progress; a DPA is available on request. - [Blog](https://yagni.app/blog): Notes from the founder on proactive engineering teams, coding agents that know the business, and earned autonomy. ## How it works - A /go run is one engineer delegating one task: map, plan, build, adversarial review, pull request, with the plan checkpoint held by the engineer and the cost printed per step. - An advisor model sizes up each task and routes it across vetted open weight models, so routine work never pays top rates; every new model release is evaled before it serves a request. - The decision ledger records what was chosen and why; agents read it before they plan, and the record stays in the app and at the CLI when a session ends. - Agent Teams carry ongoing scopes: a Team holds Responsibilities, the one Number it answers for, and the Playbook it has learned, while the workers engaged on it carry the execution. Consequential calls stage a Decision; everything leaves a Receipt. - Autonomy is earned per Team on the record: Training (nothing ships without you), Supervised (reversible work ships with an undo window), Autonomous (you review outcomes). Promotion is a button a person clicks, and the floor on irreversible actions never moves. ## Integrations YAGNI reads the tools you already pay for. It does not replace them. Native, first-party connectors today: GitHub (App installation, workspace-wide), Slack, Jira, Confluence, Linear, Google Workspace (Gmail, Calendar, Drive), Notion, HubSpot, and Stripe, plus MCP Server, Custom Service, and Database connectors for the tools you run yourself. The list grows in the order customers run on. ## Academy The YAGNI Academy is the guided way to learn the system: short lessons grouped into three tracks, from the onboarding call and your first /go run to directing agent Teams with earned autonomy. - [Academy index](https://yagni.app/academy): All courses and lessons. - [What is YAGNI and how does it work?](https://yagni.app/academy/what-is-yagni): Orientation; coding agents that know your company, the Teams, and the approval gates. - [How do you get your first Brief?](https://yagni.app/academy/your-first-brief): From the onboarding call and the GitHub App install to a Brief with real work in it. - [What is YAGNI Code and how does a /go run work?](https://yagni.app/academy/yagni-code): The CLI and desktop agent, the /go pipeline from repo map to PR, routing, and the install. - [How does the Feed work in YAGNI?](https://yagni.app/academy/working-the-feed): Staged Decisions, one item one chat, and Receipts proved from the source. - [How does YAGNI triage your email?](https://yagni.app/academy/inbox-triage): Two honest states, needs you or handled, and how to delegate the lane. - [How does YAGNI track Work and prepare your Calendar?](https://yagni.app/academy/calendar-and-work): The record, by-Team lanes, case files, and meeting preparation. - [What is a YAGNI Team and what are its Responsibilities?](https://yagni.app/academy/teams): The unit that owns an outcome, its Number, and the workers engaged on it. - [How does a Team's Playbook learn your rules?](https://yagni.app/academy/playbook): Shared rules learned from real work, Plays, and the Ladder from Training to Autonomous. - [How do you ask @yagni for help?](https://yagni.app/academy/asking-yagni): The composer, the contextual rail, the full-page Ask, and what grounds the answers. - [Which tools does YAGNI connect to, and what does it do with them?](https://yagni.app/academy/connections): The native connectors, engineering-first, and the additive rule. ## Resources - [Privacy policy](https://yagni.app/privacy) - [Terms of service](https://yagni.app/terms) ## Optional - [llms-full.txt](https://yagni.app/llms-full.txt): Extended canonical description with the full harness mechanics, the org catalog, the surface map, pricing, and security.