bug: SeaweedFS Image not found on GHCR → ImagePullBackOff

[planetf1]

Pre-requisities

• I am using the newest version of the platform (agentstack version shows that CLI and platform are the same version, and there is no newer version available)

Describe the bug
SeaweedFS fails to start: the chart-referenced image ghcr.io/i-am-bee/chrislusf/seaweedfs:4.03 returns manifest unknown, causing ImagePullBackOff and leaving object storage unavailable.

To Reproduce

1. Set seaweedfs.enabled: true and deploy AgentStack.
2. Wait for the SeaweedFS pod image pull.
3. Check pod Events → see manifest unknown and ImagePullBackOff.

Expected behavior
SeaweedFS image pulls successfully and the pod becomes Ready, providing S3-compatible storage.

Logs / Screenshots / Code snippets

Failed to pull image "ghcr.io/i-am-bee/chrislusf/seaweedfs:4.03": ... manifest unknown
Error: ImagePullBackOff

If agent is not working, please send output of the following commands:

agentstack list
# If some of the agents are in error state
agentstack logs <agent-id>
agentstack info <agent-id>

Set-up:

• Model provider: N/A

Additional context

• Docs: No explicit requirement found to login to GHCR for SeaweedFS. For private registries, docs recommend Kubernetes imagePullSecrets (see docs/development/deploy-agent-stack/deployment-guide.mdx → "Private image registries").
• Error type (manifest unknown) suggests a missing/non-existent tag/repository rather than an authentication issue.
• Chart values note a subchart bug near the SeaweedFS image repo (helm/values.yaml comment). Please verify/publish the intended image/repo/tag or document an override.

[dosubot]

The root cause is a misconfiguration in the Helm chart: the SeaweedFS image repository is set to i-am-bee/chrislusf instead of the correct i-am-bee/chrislusf/seaweedfs, and the image tag is not explicitly set. This leads to Kubernetes trying to pull ghcr.io/i-am-bee/chrislusf/seaweedfs:4.03, which does not exist, resulting in the manifest unknown and ImagePullBackOff errors. There's a comment in the chart noting this subchart bug, but it hasn't been resolved yet (source).

To work around this, you can override the image repository and tag in your Helm values. For example, in your values.yaml or with --set flags:

seaweedfs:
  image:
    registry: ghcr.io
    repository: chrislusf/seaweedfs
    tag: 4.03

Or via CLI:

helm upgrade --install agentstack ./helm \
  --set seaweedfs.image.registry=ghcr.io \
  --set seaweedfs.image.repository=chrislusf/seaweedfs \
  --set seaweedfs.image.tag=4.03

Alternatively, you can use an external S3-compatible storage by disabling SeaweedFS and configuring the external storage settings.

This is a known bug in the subchart, and the maintainers will need to fix the repository path and ensure the correct image is published to GHCR for a permanent solution.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}