Guides & docs
Heygents is an AI-powered control center for the projects you actually run. It gives every project a single workspace with live health status, a Deep Scan that reads the real repository and returns a prioritized task list, streaming runs of Claude, Codex or Gemini against the real code, and scheduled automations. These pages document every screen and collect practical guides for solo developers.
Practical, no-nonsense guides for solo developers and small teams - how to manage projects, stay organized and ship without team-sized overhead.
All 12 guides
The Solo Developer's Weekly Review Routine
A repeatable 30-minute weekly review for solo developers: empty the inbox, review what shipped, re-rank the backlog and plan next week. A Friday checklist.
Updated
Auto-Fixing Lint Errors, Broken Tests, and Missing Dependencies
Which build failures are safe to fix automatically - lint, formatting, tests broken by a known change - and the diff-review loop that clears them hands-free.
Updated
Autonomous Coding Agents Explained: What They Are and When to Trust Them
What an autonomous coding agent is, the spec-to-verified-report loop it runs, its levels of autonomy, and where a human gate is still required to stay safe.
Updated
Continuous Testing on Every Commit: A Solo Developer's Guide
Run your tests automatically on every change, get a readable failure report in under a minute, and keep the suite fast. Continuous testing for solo devs.
Updated
From Idea to Verified Commit in Under 10 Minutes
The end-to-end loop that turns a small task into a verified commit: scope, spec, code, tests, verification and a report, in under 10 minutes when scoped.
Updated
How to Keep a Test Suite Fast Enough to Run on Every Change
Keep your test suite fast enough to run on every change: the test pyramid, parallel runs, affected-only tests, isolated e2e tests and a hard time budget.
Updated
How to Manage Projects as a Solo Developer
A practical method for managing software projects as a team of one: capture work, prioritize a backlog, run solo sprints and avoid burnout. Start today.
Updated
How to Write a Task Spec an Agent Can Actually Execute
How to write a task spec an AI coding agent can run end to end: goal, scope boundaries, acceptance criteria, constraints, and done defined as green checks.
Updated
One Screen Per Project: The Live Project Dashboard
Why one live screen per project - build, tests, git state, running services - beats a terminal and a wall of tabs, and what belongs on that dashboard.
Updated
Reviewing AI-Generated Diffs Without Slowing Down
A fast review process for agent-proposed diffs: what to always check, when to trust automatic fixes, and how to keep a light human gate without scope creep.
Updated
Test Failure Reports You Can Actually Read
What makes a test failure report readable: the one failed assertion, expected vs. actual, the shortest repro, and how to tell a flaky failure from a real one.
Updated
Zero-Regression Deploys: Block Shipping Until Every Check Passes
Ship without regressions as a solo developer: make tests, lint, type checks and smoke checks a hard gate before every deploy, keep it fast, skip rollbacks.
Updated
Frequently asked questions
What is Heygents?
Heygents is an AI-powered control center for a developer's projects. Each project gets a single workspace that holds its live status, a plan with the goal and roadmap, a task list produced by Deep Scan reading the real repository, streaming AI runs against the real working directory, and scheduled automations that scan, fix or health-check on their own.
Who is Heygents for?
Solo developers and small independent makers who run a handful of their own projects - a few side projects and maybe one or two with real users. The typical user is a capable developer but not an infrastructure specialist, and wants their projects watched, checked and improved without becoming a DevOps engineer or adding team-sized process.
What is Deep Scan?
Deep Scan is a read-only AI review that opens the project's real working directory and reads its code, configuration, dependencies and logs, then returns a short list of concrete tasks. Each task has a category, an urgency, an explanation of why it matters, and a ready-to-paste instruction. Scans are grounded in the project's goal, so the suggestions fit the product.
Do I need an API key to use Heygents?
No. AI runs use the logins of the command-line tools already installed on your machine, meaning they lean on an existing subscription rather than a pay-per-use key. The API key vault in Settings stores credentials for explicit, user-initiated uses, and provider keys are removed from every AI process Heygents launches.
Where do I start?
Read Introduction & setup for signing in, the app shell and creating your first project. Then Projects & workspace for the tabs, and Tasks & Deep Scan for how the task list is produced. The Guides section covers the practices around all of this, from a weekly review routine to writing task specs an agent can execute.
Turn these guides into real work
Heygents gives every project a board, a backlog, and an AI assistant that helps you plan and ship - built for solo developers and small teams.
Open Heygents →