CP-01 · Service
One cache in front of every package registry
Package Proxy sits between your builds and the public registries. The first request for a package goes upstream; every request after that is served from a cache in the same region as your runners, with integrity checks intact.
- Registries
- 18, from npm and PyPI to apt and apk
- Endpoint
- https://acme.cacheport.cc/<registry>/
- Regions
- fra1, iad1, sjc1, sin1, syd1
- Metadata
- Refreshed every 5 min, stale copy on upstream failure
- Artifacts
- Immutable, stored per version
- Auth
- Scoped read tokens, CI-friendly
What you get
Everything in Package Proxy
Read-through caching
Nothing to pre-sync. Packages are fetched on first use and kept, so the cache matches what your builds actually need.
Upstream outage protection
If a registry slows down or goes offline, cached versions keep serving and metadata falls back to the last good copy.
Rate limit relief
Hundreds of runners look like one well-behaved client to the upstream, so shared NAT addresses stop tripping per-IP limits.
Release cooldown
Hold back versions younger than a set age. Fresh installs pick the newest version that has been public long enough to be vetted.
Allow and deny rules
Block a package, a version range or a whole scope across every repository at once, with the reason shown to whoever hits it.
Private upstreams
Add GitHub Packages, GitLab, Artifactory or your own registry behind the same URL. Private names never leak to public registries.
Audit log
Every pull is recorded with token, repository, package, version and outcome. Export to S3 or stream to your SIEM.
Cache-Status headers
Each response says whether it came from the cache, from upstream or from a stale copy, using the standard Cache-Status header.
Start
Up and running in three steps
Create a workspace
Sign up, pick the region closest to your runners and create a read token for CI.
terminal $ cacheport login $ cacheport workspace create acme --region fra1 $ cacheport token create ci --scope readPoint your clients at it
Each package manager needs one setting. Store the token as a CI secret.
config registry=https://acme.cacheport.cc/npm/ index-url = https://acme.cacheport.cc/pypi/simple/ GOPROXY=https://acme.cacheport.cc/go,directRun your pipeline
The first run warms the cache. From the second run on, the pull log shows hits.
ci log $ npm ci added 1214 packages in 9s # Cache-Status: cacheport; hit
| Supported registries | npm, Yarn, pnpm, PyPI (pip, Poetry, uv), Conda, Maven, Gradle, Go modules, Cargo, RubyGems, Composer, NuGet, Docker Hub, Helm, apt, Alpine apk |
|---|---|
| Regions | Frankfurt (fra1), Ashburn (iad1), San Jose (sjc1), Singapore (sin1), Sydney (syd1) |
| Protocol | HTTPS only, TLS 1.2 and 1.3, HTTP/2 |
| Metadata freshness | 5 minutes by default, configurable from 1 minute to 24 hours per registry |
| Largest single artifact | 10 GB |
| Authentication | Bearer tokens, Basic auth for clients that need it, OIDC from GitHub Actions and GitLab CI |
| Availability | 99.9% on Team, 99.95% on Business, custom on Enterprise |
Ecosystems
Setup guides by package manager
Common uses
Where it pays off first
Monorepos with thousands of dependencies
Cold installs that took minutes come down to the time it takes to copy files from a nearby cache.
Self-hosted runner fleets
Runners behind one NAT address stop competing for the same per-IP limits on public registries.
Regulated builds
Every dependency a release used is recorded, kept and reproducible, even if it is later removed upstream.
FAQ
Questions
Is a package cached for everyone in my workspace?
Yes. A workspace has one shared cache. The first repository to pull a version warms it for every other repository and runner.
What happens if a package is unpublished upstream?
Versions already in your cache stay available, so existing lockfiles keep installing. You can block a version yourself with a deny rule if you want it gone.
Does Cacheport modify packages?
Never. Artifacts are stored and served byte for byte, so checksums in lockfiles and go.sum files keep matching.
Can I use it from developer laptops too?
Yes. Create a personal token and use the same settings locally. Many teams do this to get the same dependency set in CI and on laptops.
Works with the rest of Cacheport
Same workspace, same tokens
Start on the free plan
20 GB of cache, every registry, unlimited users. Upgrade when your pipelines outgrow it.