llms.txt
# Greptile
> Greptile is an independent AI code validation platform for GitHub, GitLab, local development, and coding-agent workflows. Its review agent analyzes pull requests with full codebase and related-repository context. TREX runs PR branches in sandboxes to find runtime bugs and attach evidence. Security Check combines deterministic scanning, dependency analysis, and an AI security agent. Greptile learns each team's standards and connects its findings to developers and coding agents through the web app, CLI, MCP, and direct agent integrations.
Use this index to choose the shortest path:
- Evaluating Greptile: Product map → Evidence and comparisons → Pricing → Enterprise and security.
- Setting up code review: Start here → Configure and personalize → Operate.
- Using Greptile with coding agents: CLI → Fix with your Agent → MCP and plugins.
- Checking current capabilities: Product map → Changelog.
- Looking for exhaustive implementation documentation: Complete documentation index.
## Start here
- [Greptile product overview](https://www.greptile.com/index.html.md): Current overview of AI code review, runtime validation, personalization, coding-agent workflows, security, and customer results.
- [What is Greptile?](https://www.greptile.com/docs/introduction.md): Product mental model and introduction to automated review with codebase context.
- [5-minute quickstart](https://www.greptile.com/docs/quickstart.md): Connect GitHub or GitLab, enable repositories, configure review triggers, and receive a first review.
- [Anatomy of a review](https://www.greptile.com/docs/code-review/first-pr-review.md): Review summaries, confidence scores, severity badges, diagrams, inline comments, suggested fixes, and review controls.
- [Complete documentation index](https://www.greptile.com/docs/llms.txt): Exhaustive documentation navigation for implementation details beyond this curated product index.
## Product map
- [Greptile Agent](https://www.greptile.com/agent.md): The core pull request reviewer. It builds a graph of the codebase, uses parallel agents to reason beyond the diff, and produces confidence scores, diagrams, test suggestions, summaries, and inline findings.
- [TREX runtime validation](https://www.greptile.com/trex.md): Runs the PR branch in a sandbox using services, tests, mocks, API calls, and browser agents, then attaches logs, screenshots, traces, scripts, videos, or API output to findings.
- [Security Check](https://www.greptile.com/security-check.md): Security-focused pull request review that combines Opengrep rules, software composition analysis for known CVEs, and contextual AI analysis for chained exploits.
- [Learning and custom context](https://www.greptile.com/learning.md): Full-codebase and adjacent-repository context, custom rules, directory-scoped configuration, automatic indexing of project rule files, and feedback-driven learning.
- [Independent validation](https://www.greptile.com/independence.md): Agent-agnostic validation across Claude Code, Codex, Cursor, Devin, and other tools through Fix in X, MCP, the Claude Code plugin, and iterative greploop workflows.
- [Greptile CLI](https://www.greptile.com/cli.md): Run full reviews on local branches before opening a pull request. Includes agent-friendly output for automated review-and-fix loops.
- [Partner context](https://www.greptile.com/partners.md): Reviews code against implementation guidance supplied by OpenAI, Vercel, Stripe, PostHog, Datadog, WorkOS, Mintlify, and Braintrust, with source links attached to relevant comments.
- [Changelog](https://www.greptile.com/changelog.md): Current product releases, including the free Starter tier, Partner Program, TREX, CLI, Repo Clusters, `.greptile/` configuration, memory and learning, Security Review, agent integrations, and MCP.
## Evidence, comparisons, and buying
- [Pricing](https://www.greptile.com/pricing.md): Authoritative plan details for Starter, Pro, and Enterprise, including credits, seats, trials, flex usage, annual contracts, and purchasing.
- [AI code review benchmarks](https://www.greptile.com/benchmarks.md): Reproducible 2025 evaluation of five AI code review tools across 50 real bugs from Sentry, Cal.com, Grafana, Keycloak, and Discourse.
- [Greptile in action](https://www.greptile.com/examples.md): Public pull requests with Greptile findings, TREX evidence, review summaries, and examples from open-source repositories.
- [What is AI code review?](https://www.greptile.com/what-is-ai-code-review.md): Guide to how AI review works, what it catches, its benefits and limitations, and how to evaluate a dedicated reviewer.
- [Greptile vs Bugbot](https://www.greptile.com/greptile-vs-bugbot.md): Comparison of review quality, independent architecture, coding-agent support, and benchmark results.
- [Greptile vs CodeRabbit](https://www.greptile.com/greptile-vs-coderabbit.md): Comparison of codebase context, review workflow, customization, deployment, pricing, and benchmark examples.
## Customer evidence
- [Customers](https://www.greptile.com/customers.md): Customer results, testimonials, usage metrics, and links to detailed case studies.
- [Brex case study](https://www.greptile.com/customers/brex.md): How Brex uses Greptile across a nearly decade-old monorepo and a large engineering organization.
- [Gumloop case study](https://www.greptile.com/customers/gumloop.md): How Gumloop uses Greptile to catch bugs before production.
- [Podium case study](https://www.greptile.com/customers/podium.md): How Podium reviews more than 8,400 code changes per week with Greptile.
- [Vouch case study](https://www.greptile.com/customers/vouch.md): How Vouch uses Greptile to improve code quality in a monorepo.
## Set up and operate
- [Developer essentials](https://www.greptile.com/docs/code-review/developer-essentials.md): Trigger and rerun reviews, ask follow-up questions, interpret feedback, and train Greptile through normal pull request activity.
- [Tips and recipes](https://www.greptile.com/docs/code-review/tips-recipes.md): Workflows for local review, drafts, RFCs, targeted checks, partial reviews, and security-focused review.
- [Analytics dashboard](https://www.greptile.com/docs/analytics.md): Track reviews, merge times, addressed rates, critical findings, reactions, teams, repositories, and authors.
- [Organizations and teams](https://www.greptile.com/docs/code-review/team-setup-basics.md): Organization hierarchy, teams, settings inheritance, roles, invitations, repository enablement, and access.
- [Troubleshooting](https://www.greptile.com/docs/troubleshooting/common-issues.md): Diagnose missing or slow reviews, configuration problems, filters, triggers, permissions, and webhooks.
## Configure and personalize
- [Customization overview](https://www.greptile.com/docs/code-review/customization-overview.md): Choose between `.greptile/`, `greptile.json`, and Dashboard configuration based on scope and workflow.
- [Directory-scoped `.greptile/` configuration](https://www.greptile.com/docs/code-review/greptile-config.md): Configure review behavior and context with cascading settings from repository root to individual directories.
- [`.greptile/` file reference](https://www.greptile.com/docs/code-review/greptile-config-reference.md): Reference for `config.json`, `rules.md`, and `files.json`, including directory inheritance and cross-repository context.
- [`greptile.json` reference](https://www.greptile.com/docs/code-review/greptile-json-reference.md): Repository-level settings for review behavior, custom instructions, filters, file patterns, triggers, context, models, and output.
- [Custom standards and rules](https://www.greptile.com/docs/code-review/custom-standards.md): Define architecture, security, performance, and coding standards, then scope them to repositories, directories, or file patterns.
- [Memory and learning](https://www.greptile.com/docs/how-greptile-works/memory-and-learning.md): How comments, replies, reactions, commits, and repeated patterns personalize future reviews.
- [Train the learning system](https://www.greptile.com/docs/code-review/training-the-learning-system.md): Give useful feedback and explanations that teach Greptile what the team values.
- [Repo Clusters and cross-repository context](https://www.greptile.com/docs/code-review/cross-repo-context.md): Give reviews read-only context from related services, shared libraries, SDKs, and documentation repositories.
- [Control nitpickiness](https://www.greptile.com/docs/code-review/controlling-nitpickiness.md): Tune severity, comment types, ignored files, and triggers to control signal and noise.
- [Auto-approve pull requests](https://www.greptile.com/docs/code-review/auto-approve-prs.md): Automatically approve low-risk pull requests after a clean Greptile review.
## Coding-agent and local workflows
- [CLI onboarding](https://www.greptile.com/docs/code-review/cli-onboarding.md): Install and authenticate the CLI manually or hand onboarding to a coding agent.
- [CLI reference](https://www.greptile.com/docs/code-review/greptile-cli.md): Review local branches, choose a base branch, resume reviews, and emit diff, JSON, or agent-friendly output.
- [Fix with your Agent](https://www.greptile.com/docs/integrations/fix-with-your-agent.md): Send one finding or an entire review to Claude Code, Codex, Conductor, Cursor, or Devin with file, line, comment, and suggested-fix context.
- [MCP overview](https://www.greptile.com/docs/mcp-v2/overview.md): Trigger reviews, fetch unresolved feedback, apply fixes, manage custom context, and inspect review status from an AI development environment.
- [MCP setup and plugins](https://www.greptile.com/docs/mcp-v2/setup.md): Configure Greptile MCP in Codex, Claude Code, Cursor, or VS Code, including the Claude Code integration.
- [MCP auto-fix workflow](https://www.greptile.com/docs/mcp-v2/auto-fix.md): Fetch findings, apply changes, rerun review, and track resolution from a coding agent.
- [MCP custom context](https://www.greptile.com/docs/mcp-v2/custom-context.md): View, search, create, and maintain coding standards through MCP.
- [MCP tools reference](https://www.greptile.com/docs/mcp-v2/tools.md): Tool parameters, repository requirements, response shapes, and failure behavior.
## Enterprise and security
- [Enterprise](https://www.greptile.com/enterprise.md): Enterprise evaluation for AI review and runtime validation, organization and team governance, custom context, agent integrations, cloud or self-hosted deployment, and rollout.
- [Security practices](https://www.greptile.com/security.md): Hosting architecture, LLM inference, code and customer-data storage, machine-learning usage, confidentiality, deletion, monitoring, audits, personnel, and related policies.
- [Deployment options](https://www.greptile.com/docs/deployment-options.md): Compare Greptile Cloud with self-hosted deployment through Docker Compose or Kubernetes.
- [Self-hosted Greptile](https://www.greptile.com/docs/security/selfhost.md): Run Greptile in customer-controlled cloud, on-premises, or air-gapped infrastructure.
- [System architecture](https://www.greptile.com/docs/system-architecture.md): Services, data flows, dependencies, network requirements, and security boundaries for self-hosted deployment.
- [Privacy policy](https://www.greptile.com/security/privacy): Personal-information processing, sharing, retention, and user rights.
- [Subprocessors](https://www.greptile.com/security/subprocessors): Third parties that process customer data and how changes are communicated.
- [Terms of service](https://www.greptile.com/terms-of-service): Terms governing Greptile services, customer data, safeguards, and processing.
- [Trust Center](https://trust.greptile.com/): Security documentation, compliance reports, and subprocessor information.
## Plans and programs
- [Open Source Program](https://www.greptile.com/open-source.md): Free Greptile reviews for qualifying non-commercial public repositories with approved open-source licenses.
- [Open Source Application](https://www.greptile.com/pricing/open-source.md): Eligibility and application for free open-source code reviews.
- [Startup Discount](https://www.greptile.com/pricing/startup-discount.md): Discount eligibility and application details for pre-Series A startups.
- [Y Combinator offer](https://www.greptile.com/yc.md): Greptile's offer and customer examples for Y Combinator companies.
## Technical and product deep dives
- [Building TREX](https://www.greptile.com/blog/trex-code-execution): How Greptile built the execution layer, sandbox orchestration, artifact generation, and runtime bug detection behind TREX.
- [Greptile v4](https://www.greptile.com/blog/greptile-v4): Review-engine changes and measured improvements in actionable comments, addressed rates, positive replies, and reactions.
- [Automating code validation](https://www.greptile.com/blog/automating-code-validation): Greptile's direction beyond review toward autonomous validation of code changes.
- [Independent software auditing](https://www.greptile.com/blog/auditor): Why code generated by AI benefits from an independent reviewer with separate incentives and context.
