Kodus is an open-core AI code-review agent whose reviewer, Kody, comments on pull requests with repository-wide context and enforces team standards written as plain-language Kody Rules. It suits engineering teams whose pull requests wait on human reviewers and who want business context from tools such as Jira, Linear, and Notion applied automatically.[1][2]
Quick verdict
Teams whose pull requests stall waiting for human review will get the most from Kodus. Kody watches GitHub, GitLab, Bitbucket, Azure DevOps, and Forgejo or Gitea pull requests, reads across files and the wider repository, and checks changes against Kody Rules plus synced business context. The tradeoff is setup and cost shape: developers still own repository connections, rule authoring, and self-hosting work, while the managed Teams plan bills per developer plus token usage. If your team merges daily across those git platforms and a developer can own the rules, choose Kodus; if you want review with no configuration or no per-seat AI bill, look at a managed reviewer first.[1][2][3]
Key features
Cross-file pull-request reviews: Kody reads across files and the wider repository so feedback reflects how a change interacts with surrounding code, covering security, performance, best practices, and clean code.[1][2]
Plain-language Kody Rules: team standards live in .kody/rules files written in ordinary language, so conventions are enforced without building a custom linter.[1][2]
Business-rule validation: syncing Jira, Linear, Notion, Figma, and Slack lets Kody check implementation against the documented requirement, with custom business rules on Enterprise.[1]
Conversational review chat: developers ask follow-up questions inside the pull request and Kody answers with conversation memory, keeping discussion attached to the code.[1]
Kody Issues tech-debt tracking: findings worth fixing later become tracked Kody Issues with configurable severity levels, beyond the pull-request thread.[1]
Cockpit delivery metrics: a dashboard reports review coverage and engineering delivery signals; Cockpit ships on Teams and above, not Community.[1]
Model-agnostic BYOK setup: teams bring their own keys from major providers such as OpenAI, Anthropic, and Google; Enterprise supports fully custom models.[1]
Self-hosted runners and CLI: self-hosted runners on Teams and above keep code in team infrastructure, with a CLI plus CI/CD integration and managed cloud regions in the US and EU.[1][2][3]
Who will love it
Teams drowning in pull-request queues: Kody reviews every pull request automatically, so authors get security, performance, and style feedback before a human opens the diff.[1]
Organizations that want rules without linter projects: Kody Rules encode conventions in plain language, letting leads change review behavior by editing text instead of shipping tooling.[1][2]
Teams needing self-hosted review for data control: the free self-hosted Community core plus self-hosted runners keeps review inside team infrastructure under an AGPL-3.0 core.[1][2]
Who should avoid it
Teams wanting zero-configuration review: developers still connect repositories, author and maintain Kody Rules, and tune out noise, so real setup and upkeep remain.[1]
Cost-sensitive teams wary of usage billing: Teams costs 10 dollars per developer per month, or 8 dollars annually, plus token usage that is unstated, so heavy repositories cannot budget the total from the pricing page alone.[1]
Organizations avoiding strong copyleft: the core is AGPL-3.0 with commercial enterprise components, so hosted or distributed use carries obligations beyond permissive licences.[1]
Alternatives and compatibility
Kodus is defined as an AI code-review agent: Kody reviews pull requests with cross-file and repository context across GitHub, GitLab, Bitbucket, Azure DevOps, and Forgejo or Gitea, from an open-source core created in March 2025.[1][2]
If your team merges daily across the five supported git platforms and a developer can own Kody Rules, choose Kodus; if you need review with no rule authoring, predictable per-seat pricing with no token meter, or a permissively licensed core, evaluate a managed reviewer or simpler bot first.[1][2]
Questions worth asking
What is the best Kodus alternative for non-technical users?
No substitutable catalogue alternative is named yet, because no published catalogue record shares this AI-review workflow with supporting evidence. Non-technical stakeholders do not configure Kodus directly; developers own setup while stakeholders read Cockpit metrics and Kody Issues.[1]
Is Kodus really no-code?
How much does Kodus cost per month?
Community self-hosting is free with unlimited repositories and reviewers. Teams costs 10 dollars per developer per month billed monthly, or 8 dollars billed annually, plus token usage that the pricing page does not quantify. Enterprise is custom-priced with SSO/MFA, advanced security review, and a dedicated success manager. Prices checked October 11, 2026.[1]
Sources and editorial review
Source research. Reviewed on October 11, 2026. Written by Sanjib Sah. AI assisted; reviewed by a human editor.
We have not run or tested this product.
- Kodus official site
official · Observed October 11, 2026
Official site: Kody agent, five git platforms, cross-file context, Kody Rules, business-rule sync, chat memory, Issues, Cockpit, BYOK, CLI plus CI/CD, runners, US/EU cloud.
- Kodus pricing
official · Observed October 11, 2026
Official pricing: Community free self-hosted; Teams per-developer rates plus tokens, SSO add-on; Enterprise custom with SSO/MFA and success manager.
- kodus-ai repository
official · Observed October 11, 2026
Official repository: created March 2025, 1.4k stars, Dual AGPL-3.0 and Commercial, TypeScript and Python, Kody Rules in .kody/rules.
Compared with these products
No explicit proprietary baseline is recorded.