# FlareCode

FlareCode is a cloud control tower for a fleet of coding agents across your GitHub repositories. It takes a plain-English task, works in an isolated sandbox, runs checks, and opens a Pull Request for human review.

## Product
- [FlareCode — Run your coding fleet from anywhere](https://flarecode.sh): Run a fleet of coding agents across every repo. Hand off a task in plain English, watch the fleet, and review the PR from anywhere.
- [Pricing](https://flarecode.sh/pricing): Plans, usage allowances, BYOK, bundled inference, and cost-cap behavior.
- [Features](https://flarecode.sh/features): Async-first agents, multi-repo fleet mode, private workspaces, GitHub-native PRs, and BYOK.
- [How it works](https://flarecode.sh/how-it-works): The full agent loop: connect a repo, submit a task, review the PR.
- [Security](https://flarecode.sh/security): Sandbox isolation, GitHub permissions, egress controls, secret handling, logs, and retention.
- [FAQ](https://flarecode.sh/faq): Pre-signup questions about FlareCode, pricing, permissions, models, and review workflow.
- [Compare FlareCode](https://flarecode.sh/compare): Side-by-side comparisons of FlareCode and other coding agents (Devin, Cursor, Conductor, Codex, Claude Code).
- [FlareCode alternatives](https://flarecode.sh/alternatives): How FlareCode compares as an alternative to other coding agents.

## Documentation
- [Welcome](https://flarecode.sh/docs): Cloud control tower for coding-agent fleets. Connect repos, hand off work, and review Pull Requests from anywhere.
- [Auto-merge on green](https://flarecode.sh/docs/auto-merge): Let FlareCode merge a PR for you once CI passes — opt-in, off by default, risk-tiered, and still bounded by your repo's branch protection.
- [Best practices](https://flarecode.sh/docs/best-practices): Get the most out of FlareCode — write tasks agents can finish, pick the right mode and model, and review fast.
- [Choosing a model](https://flarecode.sh/docs/choosing-a-model): Use our bundled models, or bring your own keys. How to switch and when to.
- [Connect a repo](https://flarecode.sh/docs/connect-a-repo): Install the GitHub App per repo, create projects, and revoke access at any time.
- [costs and limits](https://flarecode.sh/docs/costs-and-limits): How projects, concurrency, monthly allowances, and per-goal cost caps protect you.
- [Data and retention](https://flarecode.sh/docs/data-and-retention): What we log, what we never log, and how long anything is kept.
- [FAQ](https://flarecode.sh/docs/faq): Quick answers to the questions that come up most.
- [Getting started](https://flarecode.sh/docs/getting-started): From signup to your first Pull Request in about five minutes.
- [GitHub permissions](https://flarecode.sh/docs/github-permissions): Exactly which GitHub OAuth and App permissions FlareCode requests, and what it never asks for.
- [Hooks](https://flarecode.sh/docs/hooks): Bring your own guardrails — block, gate, or get notified at the points an agent can't route around. Enforced server-side; nothing runs until you add one.
- [Import local agent memory](https://flarecode.sh/docs/import-memory): Bring local Claude Code, Codex, Cursor, and VS Code coding-agent context into FlareCode.
- [Interactive vs goals](https://flarecode.sh/docs/interactive-vs-goals): Two ways to work with an agent, and when to reach for each.
- [Live preview](https://flarecode.sh/docs/live-preview): See the app running while the agent works on it — a dev server in the sandbox, on a private URL, auto-detected for common frameworks.
- [Memory](https://flarecode.sh/docs/memory): How FlareCode learns your repos and your preferences over time, and how to review, approve, and manage what your agents remember.
- [Multi-repo projects](https://flarecode.sh/docs/multi-repo): Use projects and agents to run work across several repos.
- [Notifications](https://flarecode.sh/docs/notifications): How FlareCode tells you a PR is ready or a task failed.
- [Plans and billing](https://flarecode.sh/docs/plans-and-billing): The five plans, what each includes, and how billing works.
- [Security model](https://flarecode.sh/docs/security-model): Where your code runs, how it's isolated, and what we never touch.
- [Submit a task](https://flarecode.sh/docs/submit-a-task): How to describe a goal, what the agent does with it, and what you get back.
- [Troubleshooting](https://flarecode.sh/docs/troubleshooting): Common failure modes and how to get unstuck.
- [Recurring workflows (/loop) and Autopilot](https://flarecode.sh/docs/workflows): Stand up a loop that keeps shipping toward a goal — or point Autopilot at your issue backlog and let it work through it, PR by PR.

## Blog
- [designing the cost cap](https://flarecode.sh/blog/designing-the-cost-cap): A runaway agent that ignores its budget is a P0. Here's how we make sure it can't happen.
- [your code stays yours](https://flarecode.sh/blog/your-code-stays-yours): Why your workspace is durably backed up — encrypted, never resold, never used for training — and still entirely yours.
- [why flarecode](https://flarecode.sh/blog/why-flarecode): A note on why a hosted coding-agent fleet exists, and who it's for.

## Changelog
- [Autopilot: point an agent at your backlog and walk away](https://flarecode.sh/changelog/autopilot-backlog): Loops can now pull work straight from Linear or GitHub issues — one PR per item, pinging you only when they need you, with a while-you-were-away digest.
- [Loops + hooks: point a fleet at a goal and set the rules](https://flarecode.sh/changelog/autonomous-workflows): Recurring workflows that keep shipping toward a north star, and bring-your-own guardrails enforced server-side.
- [Describe a goal, get a pull request](https://flarecode.sh/changelog/autonomous-goals): Hand the agent a whole goal — not just a message. It plans, builds, reviews its own work, and opens a PR for you to merge.
- [Frontier models, your keys](https://flarecode.sh/changelog/bring-your-own-models): A fast default model is included on every plan. Bring your own API key to run Claude, GPT and others — billed at provider cost, with a spend limit you set.
- [A calmer fleet IDE](https://flarecode.sh/changelog/fleet-ide-upgrades): Land on a fleet overview, jump between agents, rename and tidy as you go, and a gentler way to wake a paused agent.
- [Agents that show their work](https://flarecode.sh/changelog/self-verifying-agent): Before calling a task done, the agent runs your tests and opens the app in a real browser — and learns each repo so the next task starts smarter.
- [Live preview that survives a cold start](https://flarecode.sh/changelog/live-preview): The preview pane finds the dev server's real port, waits for it to actually serve, and self-heals across restarts.
- [Live sessions that don't stall](https://flarecode.sh/changelog/live-sessions): Streaming recovers itself mid-turn, cold agents wake gracefully, and the composer and error states got a pass.
- [Run a whole fleet from one screen](https://flarecode.sh/changelog/multi-agent-fleet): Switch between running agents in a single tab — no reload, warm sessions, instant context.
- [Privacy and terms, expanded](https://flarecode.sh/changelog/privacy-terms-expanded): Clearer data handling, retention, and deletion — covering GDPR, CCPA, and India's DPDP.
- [Publishing you can trust](https://flarecode.sh/changelog/publishing-reliability): Managed publish now pushes the agent's real work and verifies it landed before reporting success.
- [A harder sandbox around every agent](https://flarecode.sh/changelog/sandbox-hardening): Model and git credentials never touch the box your code runs on, the agent runs unprivileged, and a red-team suite runs on every change.
- [a real docs site, and light mode](https://flarecode.sh/changelog/docs-and-themes): Full documentation with search, plus light and dark themes across the whole site.
- [workspaces survive a sleep](https://flarecode.sh/changelog/workspace-persistence): Gitignored local files — .env, local config — now persist when an agent sleeps and restores.
- [The first agents shipped to production](https://flarecode.sh/changelog/v0-ga): Hand off a task in plain English, get a pull request. The hosted agent fleet is live.
- [/numbers is live](https://flarecode.sh/changelog/numbers-live): Public stats page — agents, conversations, deployments, tokens — refreshed every minute from production.
- [Email when your agent ships](https://flarecode.sh/changelog/email-on-publish): Waitlist confirmations and agent updates now send straight from our own infrastructure.

## Company and Legal
- [About](https://flarecode.sh/about): Why FlareCode exists, who builds it, and how to reach the founder.
- [Contact](https://flarecode.sh/contact): Support, sales, security, press, and founder contact routes.
- [Manifesto](https://flarecode.sh/manifesto): The philosophy behind FlareCode: async-first coding agents for builders who stay in motion.
- [Vision — The command layer for software delivery](https://flarecode.sh/vision): Where FlareCode is going: from a coding-agent fleet to a command layer where human teams and agent fleets build software together, and what's live today.
- [Privacy Policy](https://flarecode.sh/privacy): How FlareCode collects, uses, shares, retains, and protects data.
- [Terms of Service](https://flarecode.sh/terms): Eligibility, billing, acceptable use, IP, disclaimers, liability, and dispute resolution.

## Support and Security Contacts
- Support: support@flarecode.sh
- Security: security@flarecode.sh
- Privacy: privacy@flarecode.sh
- General: hello@flarecode.sh

## Machine-Readable Indexes
- [Sitemap](https://flarecode.sh/sitemap.xml): Canonical XML sitemap for public indexable pages.
- [Robots](https://flarecode.sh/robots.txt): Crawler policy for public and private surfaces.