remote-runtime
Routes OpenHands agent-server startup, REST/WebSocket routes, remote conversations, and Docker/Apptainer/API/cloud workspace workflows.
- License 许可
- MIT
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- remote-runtime
- license
- MIT
- metadata
- { "disco-role": "operating" }
- description
- Routes OpenHands agent-server startup, REST/WebSocket routes, remote conversations, and Docker/Apptainer/API/cloud workspace workflows.
- disable-model-invocation
- true
02 资源文件 4
- 参考文档references/agent-server-cli.md860 B
- 参考文档references/troubleshooting.md1.3 KB
- 参考文档references/workspaces-and-remote-conversations.md1.3 KB
- 脚本scripts/managed_api_server.py3.6 KB
03 SKILL.md 正文
Remote Runtime
Use this sub-skill for the server/runtime layer: python -m openhands.agent_server, REST and WebSocket APIs, remote conversations, RemoteWorkspace, Docker and Apptainer workspaces, runtime API and cloud workspaces, deferred init, auth, and custom tool import paths.
What this route owns
- Agent-server CLI flags, startup, and local server helpers.
/alive,/health,/ready,/server_info,/api/init,/api/conversations, and workspace routes.- Session API keys, bootstrap init keys, and warm-pool/deferred init flows.
DockerWorkspace,DockerDevWorkspace,ApptainerWorkspace,APIRemoteWorkspace, andOpenHandsCloudWorkspace.ManagedAPIServer-style helpers for local examples.
Start here
Read references/agent-server-cli.md for CLI flags and startup behavior. Read references/workspaces-and-remote-conversations.md for workspace classes and remote conversation flow. Read references/troubleshooting.md for auth, deferred init, browser, tmux, and workspace failures.
Run scripts/managed_api_server.py when you need a local helper that starts the agent-server subprocess and waits for /health.
Typical triggers
- "Start the OpenHands agent-server locally."
- "Why does
/api/initreturn 503?" - "How do I authenticate remote conversations?"
- "How do I run a Docker or Apptainer workspace?"
- "How do I import a custom tool module into the server?"
Cross-links
- For local
Conversationconstruction, go to../agent-core/SKILL.md. - For tool registration and default tools, go to
../built-in-tools/SKILL.md. - For GitHub Actions examples that launch a server, go to
../github-automation/SKILL.md.
04 同仓库更多 Skills
passnet-triton-opt
Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.
passnet-solve
END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.
passnet-skill
PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.
passnet-pattern-fusion
Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.
passnet-orchestrate
Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.
passnet-feedback
Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).