llms.txt
# Hoplite
> Hoplite is a cloud coding-agent platform for starting, supervising, and reviewing work in GitHub-backed repositories.
## Product
- [Hoplite](https://app.hoplite.sh/): Sign in or create an account to use the web application.
- [Documentation](https://hoplite.sh/docs): Product, workspace, integration, and CLI documentation.
- [Multiplayer agent](https://hoplite.sh/product/multiplayer): Hoplite is a multiplayer coding agent: every thread is shared with your workspace, so teammates can watch a run, steer it, approve sensitive steps and pick up where someone left off.
- [Slack agent](https://hoplite.sh/product/slack): Mention Hoplite in Slack to answer questions about your code, start a coding thread in the right project, steer running work, and get the pull request link back in the same Slack thread.
- [Automations](https://hoplite.sh/product/automations): Run Hoplite coding agents on a schedule or when something happens in GitHub, Sentry, Linear, Slack or a webhook. Every run is an ordinary thread your team can open, steer and review.
- [PR Auto-fix](https://hoplite.sh/product/pr-autofix): Hoplite's Auto-fix loop watches a pull request's checks and review threads, wakes the agent in the same thread to push fixes to the same branch, and stops when everything is green.
## Blog
Every post is served as Markdown to agent clients at its own URL (send Accept: text/markdown, or use curl); the index is at https://hoplite.sh/blog and the feed at https://hoplite.sh/feed.xml.
- [Ship a cross-repository change with coding agents](https://hoplite.sh/blog/ship-a-cross-repository-change-with-coding-agents): Coordinate backend and frontend changes with compatible contracts, separate pull requests, and explicit integration checks.
- [5 MCP workflows that give coding agents useful context](https://hoplite.sh/blog/5-mcp-workflows-that-give-coding-agents-useful-context): Choose MCP tools by the missing information in a task, and give each connection a clear access boundary.
- [Measure coding-agent results beyond tokens and lines of code](https://hoplite.sh/blog/measure-coding-agent-results-beyond-tokens-and-lines-of-code): Track accepted changes, review effort, cycle time, and regressions without confusing activity with impact.
- [How to run a framework migration with coding agents](https://hoplite.sh/blog/how-to-run-a-framework-migration-with-coding-agents): Break a framework migration into reviewable batches, preserve behavior, and verify the combined application before rollout.
- [From vulnerability alert to verified patch: an agent workflow](https://hoplite.sh/blog/from-vulnerability-alert-to-verified-patch-an-agent-workflow): Give a coding agent a bounded dependency patch with explicit compatibility checks and review requirements.
- [A weekly coding agent routine for technical debt](https://hoplite.sh/blog/a-weekly-coding-agent-routine-for-technical-debt): Use a scheduled investigation and small reviewed changes to make repository maintenance manageable.
- [Keep documentation in sync with code using a scheduled agent](https://hoplite.sh/blog/keep-documentation-in-sync-with-code-using-a-scheduled-agent): Schedule a bounded documentation check that verifies changed commands and opens a reviewable pull request.
- [7 coding agent automations you can put to work this week](https://hoplite.sh/blog/7-coding-agent-automations-you-can-put-to-work-this-week): Seven bounded recipes for recurring repository work, with explicit triggers, stop conditions, and review expectations.
- [Evaluate a coding agent on 30 real backlog tickets](https://hoplite.sh/blog/evaluate-a-coding-agent-on-30-real-backlog-tickets): A practical study design for measuring accepted work, failed attempts and reviewer effort across a real backlog.
- [Turn team conventions into agent instructions and reusable skills](https://hoplite.sh/blog/turn-team-conventions-into-agent-instructions-and-reusable-skills): Put standing rules in project instructions, repeatable procedures in repository skills and task details in the prompt.
- [Hand a coding-agent task to a teammate without starting over](https://hoplite.sh/blog/hand-a-coding-agent-task-to-a-teammate-without-starting-over): Make an agent task transferable with a shared thread, a pushed branch, a draft pull request and a clear next action.
- [Bring your local agent context to the cloud without starting over](https://hoplite.sh/blog/bring-your-local-agent-context-to-the-cloud-without-starting-over): Import useful session history and skills into Hoplite, with clear boundaries around credentials and project state.
- [Move an active coding-agent task to the cloud without losing the thread](https://hoplite.sh/blog/move-an-active-coding-agent-task-to-the-cloud-without-losing-the-thread): A preflight and recovery checklist for transferring a live task, including conversation, repository state, and environment checks.
- [Hand off Claude Code or Codex to the cloud and keep the task moving](https://hoplite.sh/blog/hand-off-claude-code-or-codex-to-the-cloud-and-keep-the-task-moving): Use the Hoplite CLI to create a cloud thread from a local agent session and understand exactly what transfers.
- [Choose a coding model using the work your team actually does](https://hoplite.sh/blog/choose-a-coding-model-using-the-work-your-team-actually-does): Compare coding models on representative tasks, review effort, retries and cost before choosing a default.
- [MacBook overheating while coding agents run? Check what they started](https://hoplite.sh/blog/macbook-overheating-while-coding-agents-run-check-what-they-started): Identify the processes behind sustained heat, reduce unnecessary local work, or move the agents to cloud sandboxes so the laptop stops running builds.
- [Coding agents keep running your Mac out of memory. What to check first](https://hoplite.sh/blog/coding-agents-keep-running-your-mac-out-of-memory-what-to-check-first): Use memory pressure and process ownership to diagnose RAM exhaustion, reduce concurrency, or move the agents' browsers and test workers into cloud sandboxes.
- [Why your MacBook gets slow when you run several coding agents](https://hoplite.sh/blog/why-your-macbook-gets-slow-when-you-run-several-coding-agents): Find whether parallel agents are limited by local compute, memory, services, or a remote model, or skip the tuning and give each agent its own cloud sandbox.
- [Stop parallel coding agents fighting over ports and dev servers](https://hoplite.sh/blog/stop-parallel-coding-agents-fighting-over-ports-and-dev-servers): Separate runtime resources as well as branches so parallel coding tasks test the application they actually changed.
- [How to run parallel coding agents without conflicting changes](https://hoplite.sh/blog/how-to-run-parallel-coding-agents-without-conflicting-changes): Split agent work by dependency and ownership, isolate execution, and verify the combined result before merging.
- [Using Hoplite to build Hoplite](https://hoplite.sh/blog/using-hoplite-to-build-hoplite): What our published use of Hoplite says about delegation and a workflow for making agent work reviewable.
- [Beyond the first Sentry automation: dedupe, prioritize, and verify fixes](https://hoplite.sh/blog/beyond-the-first-sentry-automation-dedupe-prioritize-and-verify-fixes): Build a bounded error-triage workflow that checks existing work, reproduces failures, and reports uncertainty.
- [From Slack thread to pull request](https://hoplite.sh/blog/from-slack-thread-to-pull-request): Turn a Slack bug report into a bounded coding task, decide which repository it belongs to, and review the result against the original failure.
- [From Linear issue to tested pull request](https://hoplite.sh/blog/from-linear-issue-to-tested-pull-request): Turn a Linear issue into an implementation task with clear acceptance checks, decide what the agent may touch in the tracker, and review the result against the ticket.
- [What should a cloud coding agent be allowed to access?](https://hoplite.sh/blog/what-should-a-cloud-coding-agent-be-allowed-to-access): Understand Hoplite's source-control boundary, sandbox secrets, tool approvals and the access decisions your team still needs to make.
- [Verify agent work with tests and a browser](https://hoplite.sh/blog/verify-agent-work-with-tests-and-a-browser): Use a reproducible UI bug to connect regression tests, runtime checks, and review evidence to the same code change.
- [Keep your Claude Code task moving after you close your laptop](https://hoplite.sh/blog/keep-your-claude-code-task-moving-after-you-close-your-laptop): Move the task into a cloud execution environment before your laptop sleeps, with a clear handoff and review plan.
- [What does a merged agent pull request actually cost?](https://hoplite.sh/blog/what-does-a-merged-agent-pull-request-actually-cost): Measure agent costs across inference, reported compute charges, failed attempts and human review.
- [How to review AI-generated pull requests without becoming the bottleneck](https://hoplite.sh/blog/how-to-review-ai-generated-pull-requests-without-becoming-the-bottleneck): Ask agents for focused diffs and useful verification, then return feedback precise enough for the agent to act on.
- [The agent-ready repository checklist: setup, secrets, tests and previews](https://hoplite.sh/blog/the-agent-ready-repository-checklist-setup-secrets-tests-and-previews): Configure a repository so an agent can install it, run the app and verify its changes in a fresh Hoplite sandbox.
- [What to delegate to a coding agent: 12 tasks and ready-to-use briefs](https://hoplite.sh/blog/what-to-delegate-to-a-coding-agent-12-tasks-and-ready-to-use-briefs): Choose bounded tasks and give the agent acceptance criteria, verification steps and a clear point at which to stop.
- [Cloud coding agents: from local setup to a tested pull request](https://hoplite.sh/blog/cloud-coding-agents-from-local-setup-to-a-tested-pull-request): Set up a cloud coding-agent workflow that can run your app, verify a change and return a pull request you can review.
- [Fix Production Errors While You Sleep](https://hoplite.sh/blog/sentry-automations): A walkthrough for wiring Sentry into a Hoplite automation. New errors get picked up on a schedule, reproduced with a failing test, and fixed in a pull request you review over coffee. Takes about ten minutes to set up.
- [Announcing Automations](https://hoplite.sh/blog/automations): Agents on a schedule. Describe the work once, pick a cadence, and Hoplite runs it in a fresh sandbox every time, ending in a pull request a human reviews. Available today on every plan.
- [Introducing Hoplite](https://hoplite.sh/blog/introducing-hoplite): Coding agents can finish real work now. The way we run them hasn't caught up: one laptop, one terminal, a human watching a spinner. Hoplite moves agents into isolated cloud sandboxes and makes the pull request the interface. Backed by Y Combinator.
## AI agent access
- [MCP server](https://hoplite.sh/docs/mcp-server): Connect an MCP client to the OAuth-protected streamable HTTP endpoint at https://api.hoplite.sh/mcp.
- [MCP discovery manifest](https://hoplite.sh/.well-known/mcp.json): Read the machine-discoverable endpoint, transport, authentication, and documentation metadata.
- [OpenAPI specification](https://hoplite.sh/openapi.json): Read the reviewed Hoplite API contract, operation schemas, and required permissions.
- [API authentication](https://hoplite.sh/docs/api/authentication): Create a scoped API key and send it using X-Api-Key or Bearer authentication.
- [MCP integration guide](https://hoplite.sh/docs/integrations/mcp): Configure MCP servers that Hoplite agents can use in a project.
- [CLI](https://hoplite.sh/docs/cli): Install the published @usehoplite/cli npm package or CarbonCopyInc/tap/hoplite Homebrew formula.
- [CLI package](https://www.npmjs.com/package/@usehoplite/cli): Run the official client with npx -y @usehoplite/cli.
## When to use Hoplite
Use Hoplite when a software task should run in an isolated cloud workspace and produce reviewable repository changes. It is a strong fit for implementing features, fixing bugs, investigating failures, running tests, validating a web interface in a managed preview, and opening or updating a pull request. Use the MCP server to start or inspect authorized projects and threads from another agent; use the CLI to hand off selected local Claude Code, Codex, or OpenCode sessions. Do not use public discovery endpoints for secrets or private source code.
## Company and trust
- [About Hoplite](https://hoplite.sh/about): Learn who operates Hoplite and how the product is built.
- [Hoplite team](https://hoplite.sh/team): Meet the people building Hoplite.
- [Contact Hoplite](https://hoplite.sh/contact): Find support, sales, privacy, security, and postal contact details.
- [Privacy policy](https://hoplite.sh/privacy): Review how Hoplite handles personal data.
## Important constraints
- The web app and MCP server require an authenticated Hoplite account and workspace authorization.
- Never submit credentials, access tokens, or private repository content through public discovery endpoints.
