feat: pass partial_archive_url to evaluation workflow for resume support

[juanmichelini]

Summary

Adds the partial_archive_url input to run-eval.yml so it can be forwarded to the evaluation repo's eval-job.yml workflow. This enables resuming evaluation jobs from a previous run's archive tar.gz.

Companion PR: OpenHands/evaluation#326

Changes to .github/workflows/run-eval.yml

1. New partial_archive_url workflow_dispatch input
2. Printed in the print-parameters step
3. Passed as PARTIAL_ARCHIVE_URL env var to the dispatch step
4. Included in the dispatch payload forwarded to the evaluation repo

How to test

1. Merge OpenHands/evaluation#326 first (or use eval_branch pointing to that PR's branch)
2. Trigger "Run Eval" workflow dispatch with:
   • eval_branch: openhands/resume-partial-runs-from-archive (points to the evaluation PR branch)
   • partial_archive_url: URL of a previous run's results.tar.gz
3. The evaluation job should download the archive, extract it, and resume from existing output

Checklist

• If the PR is changing/adding functionality, are there tests to reflect this?
  • Tests are in the companion PR (OpenHands/evaluation#326) — this PR is workflow config only
• If there is an example, have you run the example to make sure that it works?
• If there are instructions on how to run the code, have you followed the instructions and made sure that it works?
• If the feature is significant enough to require documentation, is there a PR open on the OpenHands/docs repository with the same branch name?
• Is the github CI passing?

---

Agent Server images for this PR

• GHCR package: https://github.com/OpenHands/agent-sdk/pkgs/container/agent-server

Variants & Base Images

Variant	Architectures	Base Image	Docs / Tags
java	amd64, arm64	eclipse-temurin:17-jdk	Link
python	amd64, arm64	nikolaik/python-nodejs:python3.13-nodejs22	Link
golang	amd64, arm64	golang:1.21-bookworm	Link

Pull (multi-arch manifest)

# Each variant is a multi-arch manifest supporting both amd64 and arm64
docker pull ghcr.io/openhands/agent-server:0365297-python

Run

docker run -it --rm \
  -p 8000:8000 \
  --name agent-server-0365297-python \
  ghcr.io/openhands/agent-server:0365297-python

All tags pushed for this build

ghcr.io/openhands/agent-server:0365297-golang-amd64
ghcr.io/openhands/agent-server:0365297-golang_tag_1.21-bookworm-amd64
ghcr.io/openhands/agent-server:0365297-golang-arm64
ghcr.io/openhands/agent-server:0365297-golang_tag_1.21-bookworm-arm64
ghcr.io/openhands/agent-server:0365297-java-amd64
ghcr.io/openhands/agent-server:0365297-eclipse-temurin_tag_17-jdk-amd64
ghcr.io/openhands/agent-server:0365297-java-arm64
ghcr.io/openhands/agent-server:0365297-eclipse-temurin_tag_17-jdk-arm64
ghcr.io/openhands/agent-server:0365297-python-amd64
ghcr.io/openhands/agent-server:0365297-nikolaik_s_python-nodejs_tag_python3.13-nodejs22-amd64
ghcr.io/openhands/agent-server:0365297-python-arm64
ghcr.io/openhands/agent-server:0365297-nikolaik_s_python-nodejs_tag_python3.13-nodejs22-arm64
ghcr.io/openhands/agent-server:0365297-golang
ghcr.io/openhands/agent-server:0365297-java
ghcr.io/openhands/agent-server:0365297-python

About Multi-Architecture Support

• Each variant tag (e.g., 0365297-python) is a multi-arch manifest supporting both amd64 and arm64
• Docker automatically pulls the correct architecture for your platform
• Individual architecture tags (e.g., 0365297-python-amd64) are also available if needed

[github-actions]

Python API breakage checks — ✅ PASSED

Result: ✅ PASSED

Action log

[github-actions]

REST API breakage checks (OpenAPI) — ✅ PASSED

Result: ✅ PASSED

Action log

[all-hands-bot]

🟢 Good taste - Clean parameter pass-through. Simple, solves a real problem (resuming expensive evals), no breaking changes. LGTM.