Skip to main content

OpenClaw agent adapter for Harbor — a standalone shim for clean upstream Harbor, until the native integration (harbor-framework/harbor#1490) merges.

Project description

harbor-openclaw

OpenClaw agent adapter for Harbor.

A standalone shim so users on the upstream pip install harbor can run the OpenClaw agent against Harbor tasks today, before the native integration (harbor-framework/harbor#1490) merges.

Install

pip install harbor harbor-openclaw

Usage

harbor run \
  -a nop \
  --agent-import-path harbor_openclaw:OpenClaw \
  -p path/to/task \
  -m openai/gpt-4o-mini

-a nop is a throwaway — Harbor requires an agent name to pass CLI validation, but --agent-import-path takes precedence and loads OpenClaw from this package instead.

Optional runtime env vars (via --ae)

All three are opt-in; omit them for a vanilla API-key run.

Var Purpose
OPENCLAW_AUTH_PROFILES_PATH Path on the host to an auth-profiles.json for OAuth/subscription providers. Base64-injected into $HOME/.openclaw/agents/<id>/agent/auth-profiles.json in the container, then scrubbed at teardown so a leaked image cannot exfiltrate the token. OPENCLAW_AUTH_B64 accepts a pre-encoded blob directly.
OPENCLAW_MEMORY_DIR Path inside the container to seed into <workspace>/memory/ before the run; openclaw memory index --force is run after seeding (best-effort — falls back to keyword search if the optional vector-index deps aren't installed).
OPENCLAW_PERSONALITY_DIR Path inside the container; files are flat-copied (by basename, -maxdepth 1) into the workspace root — useful for task-supplied AGENTS.md / persona files.

Example with all three:

harbor run \
  -a nop \
  --agent-import-path harbor_openclaw:OpenClaw \
  -p path/to/task \
  -m openai/gpt-4o-mini \
  --ae OPENCLAW_AUTH_PROFILES_PATH=$HOME/.openclaw/agents/main/agent/auth-profiles.json \
  --ae OPENCLAW_MEMORY_DIR=/app/optional_memory \
  --ae OPENCLAW_PERSONALITY_DIR=/app/persona

Migration after PR merge

Once harbor-framework/harbor#1490 lands on PyPI, you no longer need this package. Drop the --agent-import-path flag and switch -a nop to -a openclaw:

harbor run -a openclaw -p path/to/task -m openai/gpt-4o-mini

Development

uv sync --group dev
uv run pytest tests/

License

Apache-2.0

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

harbor_openclaw-0.1.0.tar.gz (183.8 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

harbor_openclaw-0.1.0-py3-none-any.whl (11.6 kB view details)

Uploaded Python 3

File details

Details for the file harbor_openclaw-0.1.0.tar.gz.

File metadata

  • Download URL: harbor_openclaw-0.1.0.tar.gz
  • Upload date:
  • Size: 183.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.9.26 {"installer":{"name":"uv","version":"0.9.26","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for harbor_openclaw-0.1.0.tar.gz
Algorithm Hash digest
SHA256 4edc08a79cf48efa844eac602991f7823715bc935e49aa1a413a35e794c5b5f5
MD5 71779d0130b68a2898c8e1ff49ed282c
BLAKE2b-256 9e2f8cc0ffc2cb85fbd22436d428a09b1f25c9233f28f6a8e06da7433b324a6e

See more details on using hashes here.

File details

Details for the file harbor_openclaw-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: harbor_openclaw-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 11.6 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.9.26 {"installer":{"name":"uv","version":"0.9.26","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for harbor_openclaw-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 306a53348003efe7e40501e7e88336ad019a5d0084fac413e0ece04c0723bc6b
MD5 93a5fa3f00162536d9efafe2e04ac076
BLAKE2b-256 415333beeca55577997acb9598f883a11c5af797a30681dfca5cc9f01732882a

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page