For teams and enterprise

Govern agent‑written code across every repository.

Your agents write more of the code every week. Repowise checks every pull request the same way, shows where health is slipping across the portfolio and shows which agent wrote what. Hosted on Teams, or inside your own network.

A finding holds the merge until the fix lands.
Every pull request shows how health moves on the files it touched.
Every repository on one grid, the one to fix first in front.

What the org gets

Every repository, read the same way by everyone.

  • Ownership and bus factor

    Who knows each file, rolled up across repositories, with bus-factor risk flagged early.

    Ownership and git intelligence
  • Contracts and blast radius

    HTTP, gRPC and topic contracts between repositories, and how far a change reaches beyond its own repo.

    Change risk across repositories
  • Rules per component

    Payments code held to a higher bar: rules in policies.yaml, next to the code they govern, checked on every change locally or in CI.

    Change policy docs
  • Reports for auditors

    Change management evidence for SOC 2, ISO 27001 and PCI DSS reviews, built from the index, each item labeled with how it is known.

    Evidence report docs
  • A shared index per repository

    Indexed once, read by every member, reviewer and agent from the same snapshot.

  • PR review and gates

    One comment per pull request, an optional merge gate, and CI gates for coverage, secrets and change risk.

  • Portfolio health

    One ranked list of what to fix this week across every repository, with drift alerts to Slack or a webhook.

  • AI provenance

    Which commits coding agents wrote, from git history alone, next to their health and owners.

  • Security rollup

    Secrets across the whole history and vulnerable dependencies your code reaches, across the portfolio.

  • Team MCP

    Every member's agent reads the same index, one URL per repository or per workspace.

One platform

One index in place of five point tools.

Docs, code health, PR review, ownership and agent context usually arrive as five tools, each reading the code its own way. Repowise reads it once: one vendor, one bill, one security review.

  • Documentation generatorsDocs and an architecture map, built from the code graph
  • Code-health scannersHealth per file, checked against real bug fixes, with a fix-first list
  • PR risk and review botsOne comment per pull request, gates in CI
  • Dependency and ownership dashboardsContracts, blast radius, owners and bus factor across repositories
  • Agent context toolsTen MCP tools, hooks and a generated CLAUDE.md for every agent

Review

Every pull request reviewed the same way.

One comment per pull request: what the change can break, files that usually change together, owners and the health change. The risk, coverage and health checks cost nothing; AI review draws from the team credit pool.

  • Optional merge gate: a check run, advisory or blocking
  • CI gates for patch coverage, new secrets and unusual change risk, on GitHub Actions or GitLab
A real AI review thread on repowise-dev/repowise #3615, as GitHub shows it: the bot flags a P1 on lazy_load.py (a generator input leaves the new Java pass with no files), says how it checked, and the author replies that commit 43ee3dc fixed it with a test.
An AI review point on repowise-dev/repowise #3615, fixed before merge.

Across repositories

See how your repositories depend on each other.

A workspace finds the HTTP, gRPC and topic contracts between repositories, the files that change together across them, and how far a change reaches beyond its own repo.

  • Workspace chat and a workspace MCP endpoint
  • Team-owned workspaces on Teams
A repowise workspace system map: seven services across three repositories, joined by HTTP, package, database and co-change edges, each with its health score and how many contracts it provides and consumes.
The system map of a workspace: services, the contracts between them and their health.

Governance

Govern the code your agents write.

See which commits coding agents wrote, from git history alone, next to health and ownership. Security triage ranks CVEs by whether your code reaches the vulnerable package, and scans secrets across the whole history.

  • PCI-DSS and SOC 2 control-coverage evidence, derived from your own findings
  • Insert-only audit trail for security data, including agent reads over MCP
Every commit labeled human, agent or unknown, from git history alone.

Integrations

It fits the tools you already run.

One org install of the GitHub App covers every member. GitLab, Bitbucket and any other CI run the same gates from the CLI. Alerts go to Slack or your own endpoint.

  • Claude Code
  • Codex
  • Cursor
  • GitHub Copilot
  • Any MCP client
  • VS Code
  • GitHub PR bot
  • GitHub Actions
  • GitLab CI
  • Bitbucket and any CI
  • Slack
  • Jira
  • Confluence

Deployment

Run it where your code lives.

Graph, git history, health, change risk, tests, dead code and PR review make no model call. Raw source is parsed in memory and never stored.

  • Self-hosted, commercialYour VPC or an air-gapped networkOnly the integrations you configure. Nothing at all in air-gapped mode.
  • Self-hosted, open sourceYour machines, with the index next to the repoOpt-out anonymous CLI telemetry, and your own model provider if you turn prose on.
  • Hosted TeamsShort-lived containers; each clone is deleted after its runYour code is read by the hosted platform, under the data handling terms.
  • Licensed the way you buy

    A commercial licence per seat, per indexed repository with unlimited seats, or enterprise-wide.

  • Support with an SLA

    A named contact, a response-time SLA and a quarterly architecture review.

  • IP indemnification

    IP indemnification and a patent grant, and no AGPL obligations on your code.

  • Built for your stack

    Custom language and framework extensions, unlimited repositories and configurable index concurrency.

SAML or OIDC single sign-on and SCIM provisioning are scoped into an enterprise rollout. Compliance reports map each control to evidence from your own findings.

FREQUENTLY ASKED

Questions, answered

How is Teams priced?

Per seat with a 3-seat minimum: $20 per seat per month, or $16 billed yearly. Each seat adds $5 a month to one shared credit pool, and the team shares 25 repositories.

What is the difference between Pro and Teams?

Pro is one person with 5 repositories and a personal balance. Teams indexes each repository once for every member and adds owner, admin and member roles, one org GitHub App install, portfolio views, team workspaces and alert webhooks.

Do we have to open-source our code because of the AGPL?

No. Using repowise inside your company, including on internal servers, creates no obligation to publish your code. A commercial licence removes the obligations that apply to modifying repowise and offering it to others.

Does the analysis send code to a model?

No. The graph, git history, code health, dead code and security scanning make no model call. Written documentation and chat use a model: on hosted repowise through our provider accounts, self-hosted through the provider and key you configure, or fully offline.

Do you support SSO and SCIM?

As part of an enterprise agreement. Tell us your identity provider and directory, and we scope SAML or OIDC single sign-on and SCIM provisioning into your rollout.

Get started

Put the whole org on one index.

Start a hosted team from 3 seats today, or talk to us about a rollout inside your network.