Repository discovery
Maps services, dependencies, conventions, boundaries, risks, and architecture, then stores the resulting context in `HYRAX.md` and `.hyrax/discovery` inside the repository.
Hyrax
Freemium
Hyrax is an AI codebase architecture and engineering review platform for teams working through GitHub. It maps repository structure, evaluates code across six engineering dimensions, and proposes verified changes as pull requests for developers to review and merge.
Hyraxとは？
Hyrax is an AI codebase architecture and engineering review platform for AI-native software teams. It reads a repository to map services, dependencies, boundaries, conventions, risks, and other architectural context, then writes that context to HYRAX.md and .hyrax/discovery so human developers and AI coding tools can work from the same reference.
The platform evaluates repositories across security, correctness, maintainability, performance, architecture, and operations. Findings are prioritized from P0 to P3 and include file and line references. Hyrax is intended to run alongside existing static scanners, AI code review tools, and dependency security products rather than replace them.
When a team chooses an issue to address, Hyrax works in an isolated worktree, applies the change, runs tests and the build, and passes the result through a 13-step verification gate. It then opens a pull request in GitHub. Developers review, approve, request changes, or dismiss findings; Hyrax does not remove the human merge decision.
Hyrax also reviews pushes against the repository's architectural context. Accepted changes add context for later cycles, while its GitHub workflow can connect upgrade work to Linear tickets and close a ticket on merge.
Hyraxでできること
Six-dimension engineering analysis
Specialized agents assess security, correctness, maintainability, performance, architecture, and operations using shared repository context.
Prioritized findings
Turns analysis into a ranked P0-to-P3 work list with file and line references, helping teams distinguish urgent issues from lower-priority guidance.
Verified upgrade pull requests
Creates changes in an isolated worktree, runs tests and the build, applies a 13-step verification gate, and returns successful work as a GitHub pull request.
Human-controlled review workflow
Developers can review, approve, dismiss, or request changes before merging. Hyrax keeps the merge decision with the engineering team.
Context that improves over time
Reviews pushes against the repository's architectural context and uses accepted changes to build more deterministic patterns for later cycles.
利用シーン
“Establish context for AI coding agents”
Teams using tools such as Cursor, Copilot, Claude Code, Codex, or Devin can give human and AI contributors a shared repository map through `HYRAX.md` and `.hyrax/discovery`.
“Find first-party issues before implementation work”
Engineering and security teams can run repository audits that inspect their own code for bugs, security risks, architectural problems, and operational concerns—not only known dependency vulnerabilities.
“Turn selected findings into reviewable changes”
A developer can choose a finding and have Hyrax prepare a tested, verified change as a pull request, while retaining the normal review and merge process.
“Maintain architecture as the codebase changes”
Teams can use push reviews and recurring context updates to detect changes against established conventions and carry accepted patterns into future work.
“Connect remediation to delivery tracking”
Organizations using Linear can connect upgrade work to tickets and close the ticket when the related pull request is merged.
よくある質問
How does Hyrax work with a repository?
After the GitHub App is installed, Hyrax performs discovery, maps the repository's anatomy and architectural context, and writes the result to `HYRAX.md` and `.hyrax/discovery`. It can then review pushes and run audits against that context.
What does Hyrax return when it finds an issue?
Findings include a priority from P0 to P3 and file and line references. For selected issues, Hyrax can create a change in an isolated worktree, verify it, and open a GitHub pull request for developer review.
Are changes merged automatically?
No. Developers review every pull request and decide whether to approve, request changes, dismiss the finding, and merge the change.
What verification happens before a pull request is opened?
Hyrax runs the tests and build in an isolated worktree and sends the change through a 13-step verification gate. Required checks can stop the run when they fail.
What is included in the Free plan?
The Free plan includes the full Hyrax product, a $30 starter credit, $10 per month in ongoing credits, and up to 100 PR reviews per month at no charge. Credits can be used for repository mapping, upgrades, reviews, and other Hyrax workflows.
クイック情報
- Category
- Developer Tool; AI Code Review
- Primary users
- Engineering leads, security teams, platform engineers, CTOs, software engineers, and DevSecOps teams
- Primary workflow
- Repository discovery → prioritized findings → verified GitHub pull requests
- Supported workflow tools
- GitHub and Linear
- Analysis dimensions
- Security, correctness, maintainability, performance, architecture, and operations
- Free plan
- Full product, $30 starter credit, $10/month ongoing credits, and up to 100 PR reviews per month free
Hyraxの代替品
Context Goblin
contextgoblin.com
Context Goblin is an AI code reviewer for GitHub pull requests that uses call-graph and cross-repository context to assess changes. It helps development teams review architecture, security, logic, and test concerns with findings verified against the codebase.
GitWarren
gitwarren.com
GitWarren is a local code-review tool for inspecting and commenting on the staged, unstaged, and untracked changes in an agent’s Git worktree before they are committed or pushed. It runs on the machine containing the repository and can be used through a desktop app, a browser, or an MCP server connected to coding agents.
Gitar
gitar.ai
Gitar is an AI-powered code review and pull request automation tool for teams that need to verify changes before merging. It reviews code, diagnoses CI failures, applies fixes under configurable rules, and checks that the pipeline passes.
Polylane
polylane.com
Polylane is an AI operations platform that connects cloud resources, code, and observability data to detect incidents, investigate causes, review production impact, and prepare fixes. It is designed for engineering and DevOps teams that want automated incident response and production change analysis.
DeepSource
deepsource.com
AI-assisted code review and static analysis for software teams.
Orca
onorca.dev
A desktop environment for developers to run and review multiple coding agents in parallel.