MetaClaw
๐ฆ Just talk to your agent โ it learns and EVOLVES ๐งฌ.
File Explorer
- logo.jpg
- metaclaw.gif
- metaclaw_mainfig.png
- metaclaw_mainfig_v2.png
- new_logo.png
- new_logo2.png
- README_AR.md
- README_DE.md
- README_ES.md
- README_FR.md
- README_HI.md
- README_IT.md
- README_JA.md
- README_KO.md
- README_PT.md
- README_RU.md
- README_VI.md
- README_ZH.md
- tinker.jpg
- video.mp4
- video_v2.mp4
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- check_backup.py
- check_done_log.py
- check_filename.py
- check_iso8601.py
- check_metadata.py
- metaclaw.json
- openclaw.json
- day01_ae8c0044-6395-42f7-8cb4-604c411defe2.jsonl
- day02_5d8c2731-eb2d-4e75-a7eb-bb31dec71523.jsonl
- day03_434e7639-a1a7-46e8-9370-22d0caae836a.jsonl
- day04_4f564f09-3237-488c-9e2f-05c18234558a.jsonl
- day05_90c70c40-40d0-4cb0-bacd-944d4f5f7f2b.jsonl
- day06_2800712c-6d8b-4232-9d76-04e97e1db347.jsonl
- day07_f4089966-4f1e-4e41-a2ff-fc75547d8784.jsonl
- day08_3257b753-1204-478a-8657-ec09ee087c28.jsonl
- day09_0203355d-1601-4162-8040-53651926872b.jsonl
- day10_3a37ef96-4f68-4fe7-82a0-4b2ec4b16cbd.jsonl
- day11_242507f8-7586-4298-a299-def88cbd1804.jsonl
- day12_67578085-c2da-46c8-a6d0-21a33c0e7577.jsonl
- day13_b975df56-74e5-4463-9df2-ca13d6beb583.jsonl
- day14_effa0e25-ff0e-4e51-9769-df9c69f3bdc2.jsonl
- day15_d17f0df1-a277-48b5-949e-552409453dde.jsonl
- day16_1744a6fe-cb81-4c59-bc6e-868ebb861b10.jsonl
- day17_ad78f0aa-de15-4465-9ed1-e8ed1c46f707.jsonl
- day18_5ce33e57-fc8d-43c2-beb3-0b5a85f8b070.jsonl
- day19_bbc436f1-4305-4075-93fc-d806c75606aa.jsonl
- day20_4a0ece6a-e1a5-4d84-99ba-46b47ac8d581.jsonl
- day21_5f0f9db4-c575-438d-8140-045658d6676b.jsonl
- day22_37c5af4e-30f5-493e-8508-c3678f80c6df.jsonl
- day23_0e8eada3-5971-40f0-abfc-c603925204ac.jsonl
- day24_ad41d5a3-cb68-49e7-b817-f4baf4b43755.jsonl
- day25_5caa2175-f53c-4848-92e1-a7b817313bc8.jsonl
- day26_0d6a6699-25f8-430e-bf90-b0e860f8575f.jsonl
- day27_bc38df86-8d2c-4fcc-823b-9cefe7750f34.jsonl
- day28_b77d28d7-f96c-42a7-8d18-58381d5e8e75.jsonl
- day29_0283bbef-995f-4a72-ae5e-447887fddf90.jsonl
- day30_f4f0d7da-b004-4ff9-a0c2-e7fed6fc2e8f.jsonl
- README.md
- sprint_tasks.json
- standup_raw.txt
- blockers.md
- milestones.json
- README.md
- api_access.log
- error_events_raw.txt
- README.md
- report_template.json
- api_spec_stub.json
- changelog_stub.md
- README.md
- archive_index_stub.json
- README.md
- sprint_review_template.md
- weekly_raw_data.json
- ci_build_raw.txt
- README.md
- test_results_raw.txt
- README.md
- review_raw_notes.txt
- api_requests_raw.csv
- README.md
- user_activity_raw.json
- README.md
- resource_availability.txt
- sprint8_progress_raw.txt
- README.md
- this_week_outputs.txt
- existing_readme.md
- README.md
- sprint_review_notes_raw.txt
- monthly_api_stats_raw.json
- README.md
- user_sessions_raw.json
- legacy_config_loader.py
- README.md
- q1_data_raw.json
- README.md
- sprint8_tasks_raw.txt
- README.md
- week2_outputs_raw.txt
- api_docs.md
- q2_timeline.json
- README.md
- sprint_review_summary.md
- test_config.json
- analysis_report.json
- data_pipeline_config.json
- monthly_metrics.json
- README.md
- user_activity.json
- api_routes.py
- deploy.sh
- log_handler.py
- prod_config.py
- README.md
- requirements.txt
- q2_project_plan.json
- README.md
- resource_allocation.json
- risk_matrix.json
- sprint8_board.json
- api_benchmark.json
- app_config.json
- ci_config.yaml
- deploy_checklist.md
- README.md
- weekly_status.json
- README.md
- roadmap_draft.md
- backlog_current.json
- done.log
- README.md
- risk_register.md
- done.log
- README.md
- slo_tracker.json
- feature_config.json
- README.md
- delivery_metrics.json
- README.md
- sprint9_board.json
- constants.py
- README.md
- requirements.txt
- api_reference.md
- project_overview.md
- README.md
- README.md
- sprint9_raw_data.json
- team_roster.json
- milestone_tracker.json
- README.md
- risk_register.json
- all_rules_compliance.json
- api_changelog.md
- project_config.json
- README.md
- sprint10_board.json
- AGENTS.md
- IDENTITY.md
- SOUL.md
- TOOLS.md
- USER.md
- all_tests.json
- all_tests_metaclaw.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- questions.json
- check_backup.py
- check_done_log.py
- check_filename.py
- check_iso8601.py
- check_metadata.py
- metaclaw.json
- openclaw.json
- day01_b5fb42b3-81b5-4a7f-8655-3909e6b9280f.jsonl
- day02_7dbbf8a4-249f-4a36-86d3-e4537b7efd7a.jsonl
- day03_e6a9e1bf-ecb0-4e5f-a37e-0c2517f2526d.jsonl
- day04_eed52abc-7aa4-4cb4-9cf8-f56ed10938d5.jsonl
- day05_b619367c-56fc-430d-8dc9-9bcd27c3e41a.jsonl
- day06_240c311d-2382-401a-837b-10a41c852325.jsonl
- day07_072a3470-2cde-4bda-b202-8c9311fc827e.jsonl
- day08_8173e78e-099d-4e47-9933-eb8d36aa9a81.jsonl
- day09_e8195cb8-6dd1-422a-aa48-9ae9264a2599.jsonl
- day10_a8029559-5ce4-4933-b3f6-e8cc001dfea2.jsonl
- day11_8ed93fd6-2dee-4472-abd1-c1bbf1314cb9.jsonl
- day12_f92cbe20-66e1-4265-bc16-c98f53336a95.jsonl
- api_spec_stub.json
- changelog_stub.md
- README.md
- sprint_tasks.json
- standup_raw.txt
- api_access.log
- archive_index_stub.json
- error_events_raw.txt
- README.md
- report_template.json
- sprint_review_template.md
- weekly_raw_data.json
- ci_build_raw.txt
- README.md
- review_raw_notes.txt
- test_results_raw.txt
- README.md
- resource_availability.txt
- sprint8_progress_raw.txt
- this_week_outputs.txt
- existing_readme.md
- legacy_config_loader.py
- monthly_api_stats_raw.json
- README.md
- sprint_review_notes_raw.txt
- user_sessions_raw.json
- q1_data_raw.json
- README.md
- sprint8_tasks_raw.txt
- week2_outputs_raw.txt
- analysis_report.json
- api_docs.md
- data_pipeline_config.json
- monthly_metrics.json
- q2_timeline.json
- README.md
- sprint_review_summary.md
- test_config.json
- user_activity.json
- api_benchmark.json
- api_routes.py
- app_config.json
- ci_config.yaml
- deploy.sh
- deploy_checklist.md
- log_handler.py
- prod_config.py
- README.md
- requirements.txt
- weekly_status.json
- backlog_current.json
- README.md
- risk_register.md
- roadmap_draft.md
- delivery_metrics.json
- README.md
- slo_tracker.json
- sprint9_board.json
- api_reference.md
- constants.py
- project_overview.md
- README.md
- requirements.txt
- all_rules_compliance.json
- api_changelog.md
- project_config.json
- README.md
- sprint10_board.json
- sprint9_raw_data.json
- team_roster.json
- AGENTS.md
- IDENTITY.md
- SOUL.md
- TOOLS.md
- USER.md
- all_tests.json
- all_tests_metaclaw.json
- CLI.md
- env.md
- scripts.md
- index.ts
- openclaw.plugin.json
- buffer-memory.yaml
- dummy.yaml
- madmax.yaml
- memory.yaml
- rl-only-memory.yaml
- rl-only.yaml
- rl.yaml
- skills-memory.yaml
- skills-only.yaml
- _env_arg_example.sh
- baseline_run.py
- buffer_memory_run.py
- build_small_subset.py
- madmax_memory_run.py
- memory_run.py
- proxy_passthrough_run.py
- proxy_run.py
- rl_only_memory_run.py
- rl_only_run.py
- rl_run.py
- skills_memory_run.py
- skills_only_run.py
- __init__.py
- all_tests_structure.py
- base.py
- basic_integrity.py
- check_cmd.py
- directory_structure.py
- file_format.py
- id_consistency.py
- questions_integrity.py
- session_format.py
- workspace_integrity.py
- __init__.py
- clean_cmd.py
- __init__.py
- infer_cmd.py
- prompts.py
- query_reader.py
- __init__.py
- report_cmd.py
- __init__.py
- run_cmd.py
- __init__.py
- scoring_cmd.py
- __init__.py
- cli.py
- utils.py
- __init__.py
- test_check.py
- test_infer.py
- test_scoring.py
- pyproject.toml
- README.md
- run_conversation_opd.py
- run_conversation_replay.py
- run_conversation_rl.py
- train.jsonl
- index.ts
- openclaw.plugin.json
- package.json
- README.md
- README.md
- conversation_skills.json
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- SKILL.md
- policy.json
- __init__.py
- candidate.py
- consolidator.py
- embeddings.py
- manager.py
- metrics.py
- models.py
- policy.py
- policy_optimizer.py
- policy_store.py
- promotion.py
- README.md
- replay.py
- retriever.py
- scope.py
- self_upgrade.py
- store.py
- telemetry.py
- upgrade_worker.py
- __init__.py
- __main__.py
- api_server.py
- auth_store.py
- bedrock_client.py
- calendar_client.py
- claw_adapter.py
- cli.py
- cli_session_store.py
- config.py
- config_store.py
- data_formatter.py
- failover.py
- idle_detector.py
- launcher.py
- log_color.py
- openclaw_env_rollout.py
- prm_scorer.py
- rollout.py
- runtime_state.py
- scheduler.py
- sdk_backend.py
- setup_wizard.py
- skill_evolver.py
- skill_manager.py
- trainer.py
- utils.py
- weaver_compat.py
- cli.d.ts
- cli.js
- cli.js.map
- slash.d.ts
- slash.js
- slash.js.map
- auto-capture.d.ts
- auto-capture.js
- auto-capture.js.map
- auto-recall.d.ts
- auto-recall.js
- auto-recall.js.map
- memory-forget.d.ts
- memory-forget.js
- memory-forget.js.map
- memory-search.d.ts
- memory-search.js
- memory-search.js.map
- memory-status.d.ts
- memory-status.js
- memory-status.js.map
- memory-store.d.ts
- memory-store.js
- memory-store.js.map
- client.d.ts
- client.js
- client.js.map
- config-schema.d.ts
- config-schema.js
- config-schema.js.map
- index.d.ts
- index.js
- index.js.map
- sidecar.d.ts
- sidecar.js
- sidecar.js.map
- types.d.ts
- types.js
- types.js.map
- HANDOFF.md
- INTEGRATION_GUIDE.md
- PUBLISH_GUIDE.md
- PUBLISH_STEPS.md
- QUICK_START.md
- __init__.py
- candidate.py
- consolidator.py
- embeddings.py
- manager.py
- metrics.py
- models.py
- policy.py
- policy_optimizer.py
- policy_store.py
- promotion.py
- replay.py
- retriever.py
- scope.py
- self_upgrade.py
- store.py
- telemetry.py
- upgrade_worker.py
- __init__.py
- config.py
- __init__.py
- __main__.py
- config.py
- server.py
- .gitignore
- pyproject.toml
- cli.ts
- slash.ts
- auto-capture.ts
- auto-recall.ts
- memory-forget.ts
- memory-search.ts
- memory-status.ts
- memory-store.ts
- client.ts
- config-schema.ts
- index.ts
- sidecar.ts
- types.ts
- .gitignore
- .npmignore
- openclaw.plugin.json
- OPENCLAW_PLUGIN_SPEC.md
- package-lock.json
- package.json
- PLAN.md
- README.md
- tsconfig.json
- bench_with_tee.sh
- metaclaw_with_tee.sh
- run_e2e_skills_memory_test.py
- run_memory_ablation.py
- run_memory_ablation_direct.py
- run_multimodel_synergy.py
- run_opd_experiment.py
- run_openclaw_tinker_opd.sh
- run_rl_experiment.py
- run_v03_benchmark.py
- run_v03_benchmark_azure.py
- test_synergy.py
- conftest.py
- test_cli.py
- test_launcher.py
- test_memory_system.py
- test_openclaw_env_rollout.py
- test_runtime_state.py
- test_sdk_backend.py
- test_setup_wizard.py
- test_utils.py
- test_v03_live_tinker.py
- .gitignore
- LICENSE
- pyproject.toml
- README.md
- requirements.txt
# Installation Guide
pythonSetup Steps
curl -LO https://github.com/aiming-lab/MetaClaw/releases/download/v0.4.0/metaclaw-plugin.zip
Download the MetaClaw OpenClaw plugin package.
unzip metaclaw-plugin.zip -d ~/.openclaw/extensions
Extract the downloaded plugin to OpenClaw extensions directory.
openclaw plugins enable metaclaw-openclaw && openclaw gateway restart
Enable the plugin and restart the gateway.
pip install -e ".[rl,evolve,scheduler]"
Install the python package with full features (RL, evolve, scheduler).
metaclaw setup
Run the interactive setup wizard to configure agents and auth.
metaclaw start
Start MetaClaw in default auto mode (skills + scheduled RL training).
Key Commands
pip install -e .
Install package in lightweight skills_only mode.
pip install -e ".[rl]"
Install package with RL training support.
metaclaw setup
Run the MetaClaw initial setup wizard.
metaclaw start
Start MetaClaw in default auto mode.
metaclaw start --mode rl
Run RL training immediately on full batch without scheduler.
metaclaw start --mode skills_only
Run with skills only and no RL.
metaclaw config wechat.enabled true
Enable WeChat integration.
# Use via CDN
jsDelivrjsDelivr serves any public GitHub repository as a CDN with zero setup. Pick a version and a file to get a ready-to-paste link and snippet.
Command Glossary
Commands referenced in this DOCs, explained below.
npx
View Details โผ
npx
This command is an alias of `npm exec`.
tldr npm exec
View documentation for the original command:
openclaw
View Details โผ
openclaw
A personal AI assistant that you run on your own devices.
Some subcommands such as `onboard`, `agent`, `doctor`, etc. have their own usage documentation.
See also: `zeroclaw`.
openclaw onboard --install-daemon
Run the onboarding wizard to set up the gateway and channels:
openclaw gateway --allow-unconfigured --port {{18789}} --verbose
Start the gateway server:
openclaw message send {{[-t|--target]}} {{+1234567890}} {{[-m|--message]}} "{{Hello}}"
Send a message to a contact:
pip install
View Details โผ
pip install
Install Python packages.
pip install {{package1 package2 ...}}
Install one or more packages:
pip install {{package1 package2 ...}} {{[-U|--upgrade]}}
Upgrade all specified packages to the latest version, installing any that are not already present:
pip install {{package}}=={{version}}
Install a specific version of a package:
unzip
View Details โผ
unzip
Extract files/directories from Zip archives.
See also: `zip`.
unzip {{path/to/archive1.zip path/to/archive2.zip ...}}
Extract all files/directories from specific archives into the current directory:
unzip {{path/to/archive1.zip path/to/archive2.zip ...}} -d {{path/to/output}}
Extract files/directories from archives to a specific path:
unzip -c {{path/to/archive1.zip path/to/archive2.zip ...}}
Extract files/directories from archives to `stdout` alongside the extracted file names:
zeroclaw service
View Details โผ
zeroclaw service
Manage the ZeroClaw background service (launchd/systemd user service).
zeroclaw service install
Install the daemon service:
zeroclaw service start
Start the daemon service:
zeroclaw service stop
Stop the daemon service:
zeroclaw
View Details โผ
zeroclaw
Fast, small, and fully autonomous AI assistant infrastructure.
Some subcommands such as `onboard`, `models`, `service`, etc have their own usage documentation.
See also: `openclaw`.
zeroclaw onboard --api-key {{api_key}} --provider {{openrouter|anthropic|openai|...}}
Initialize workspace and configuration (quick setup):
zeroclaw onboard --interactive
Run the full interactive onboarding wizard:
zeroclaw agent {{[-m|--message]}} "{{Hello, ZeroClaw!}}"
Send a single message to the AI agent:
systemctl
View Details โผ
systemctl
Control the systemd system and service manager.
Some subcommands such as `disable`, `status`, `reboot` etc. have their own usage documentation.
systemctl status
Show all running services:
systemctl --failed
List failed units:
systemctl {{start|stop|restart|reload|status}} {{unit}}
Start/Stop/Restart/Reload/Show the status of a service:
launchctl
View Details โผ
launchctl
Control Apple's `launchd` manager for launch daemons (system-wide services) and launch agents (per-user programs).
`launchd` loads XML-based `*.plist` files placed in the appropriate locations, and runs the corresponding commands according to their defined schedule.
launchctl load ~/Library/LaunchAgents/{{my_script}}.plist
Activate a user-specific agent to be loaded into `launchd` whenever the user logs in:
sudo launchctl load /Library/LaunchAgents/{{root_script}}.plist
Activate an agent which requires root privileges to run and/or should be loaded whenever any user logs in (note the absence of `~` in the path):
sudo launchctl load /Library/LaunchDaemons/{{system_daemon}}.plist
Activate a system-wide daemon to be loaded whenever the system boots up (even if no user logs in):
