Skip to content
Artifact Repository Manager
Download

artifactd / Artifact Repository Manager

Great software deserves
a clear path to release.

Publish your packages. Promote with control. See what your users experience. One self-hosted workspace for your team, CI pipelines, and AI agents.

Hosts APT · Arch · NuGet · MSI · MSIX · DMG packages Built-in MCP · Runs on Linux

Not yet supported: npm, OCI, RPM, PyPI, and Maven. Explore format coverage →

artifactd management dashboard: repository activity, publications and client feedback by channel View full screen ↗
Actual artifactd interface · illustrative demo data

Inside artifactd

Your releases. The whole picture.

Explore the workspace that connects your packages, your teams, and the people using your software.

01 / The dashboard

Know what shipped. See what’s happening.

Start with a clear view of your repository: uploads, downloads, published artifacts, and client feedback. Channel trends and recent publications put the next investigation one click away.

  • Repository activity in one view
  • Published artifacts and feedback by channel
  • Direct links to the underlying catalogue and audit
Explore the documentation →
artifactd dashboard showing repository metrics, channel trends and recent publications View full screen ↗
Actual artifactd interface · illustrative demo data

02 / Package promotion

From a new build to the right audience.

Keep versions, formats, and publication channels together. Inspect an artifact and choose where it is published as it moves through alpha, beta, testing, canary, and release.

  • One catalogue for APT, Arch, MSI, MSIX, DMG, and NuGet
  • Visible version and channel context
  • Promotion for operators, CI, and MCP agents
Explore the documentation →
artifactd artifacts catalogue with the selected build and publication controls View full screen ↗
Actual artifactd interface · illustrative demo data

03 / Users & permissions

Give every team the access it needs.

Organize users into groups, then scope grants to the package, channel, and action. Let CI publish testing builds and production clients pull releases with credentials you can rotate.

  • Users and groups in a dedicated workspace
  • Separate pull and publish permissions
  • Package and channel scopes visible at a glance
Explore access control →
artifactd grants table showing group, package, channel, pull and publish permissions View full screen ↗
Actual artifactd interface · illustrative demo data
See the users workspace
artifactd users workspace with demo users and group membership View full screen ↗
Actual artifactd interface · illustrative demo data

04 / Client Feedback

The release is only the beginning.

Turn what happens after installation into input for your next development cycle. Investigate crashes, errors, bug reports, and feature requests against the affected version, then bring that evidence into AI-assisted fixes and feature work.

  • Filter reports by kind, period, package, or channel
  • Inspect exception details and client context
  • Investigate repeated issues in an artifact’s Top problems
Explore the documentation →
artifactd Client Feedback with reports tied to a package version and report details View full screen ↗
Actual artifactd interface · illustrative demo data

AI-native

Bring real-world feedback into your next release

Use crashes, errors, and user requests to guide AI-assisted development. Client Feedback provides the evidence; built-in MCP gives your AI agent access to the catalogue and release controls that help you ship the next iteration.

  • Give your AI agent release context

    Connect your AI agent through built-in MCP to inspect packages, versions, publication channels, and catalogue status on demand. Ground release decisions in the repository’s current state.

  • Ship the next iteration from CI

    CI jobs and AI agents upload with a Bearer publish token. No admin UI click in the critical path. The management admin token cannot upload.

  • Turn user feedback into development priorities

    Investigate crashes, errors, bug reports, and feature requests in Client Feedback, with their package and version context. Bring that evidence into your AI-assisted development workflow to prioritize fixes and plan the next build.

  • Least privilege for machines

    Grants scope package, channel, and action. Give a CI identity Allow publish on one software and channel — not blanket admin. MCP uses the admin token only for operator agents.

Your next release

Bring the whole cycle together.

From your first package to your next improvement. Give your team one place to publish, promote, and learn from users.

Install artifactd on your infrastructure and connect your pipeline.