بنقرة واحدة
drclaw
// Dr. Claw workspace skill for project lookup, session inspection, TaskMaster progress, OpenClaw structured schema, and event-driven reporting
// Dr. Claw workspace skill for project lookup, session inspection, TaskMaster progress, OpenClaw structured schema, and event-driven reporting
Generate/edit images with OpenAI gpt-image-2 by default, falling back to Gemini (gemini-3.1-flash-image-preview) when OPENAI_API_KEY is unset. Supports text-to-image + image-to-image; 1K/2K/4K; use --input-image for editing, --provider to force a provider, --model to override the model.
Mandatory pre-flight compute resource check before running experiments. Detects whether local/remote GPU or compute resources are actually available. If resources are unavailable, STOPS the experiment pipeline immediately and reports to the user — preventing the model from hallucinating fake experiment results. Use when: about to run experiments, deploy training, or any GPU-intensive task.
Workflow 1.5: Bridge between idea discovery and auto review. Reads EXPERIMENT_PLAN.md, implements experiment code, deploys to GPU, collects initial results. Use when user says "实现实验", "implement experiments", "bridge", "从计划到跑实验", "deploy the plan", or has an experiment plan ready to execute.
Deploy and run ML experiments on local, remote, Vast.ai, or Modal serverless GPU. Use when user says "run experiment", "deploy to server", "跑实验", or needs to launch training jobs.
Use when a quest needs one or more follow-up runs such as ablations, robustness checks, error analysis, or failure analysis after a main experiment.
Use when a quest needs to attach, import, reproduce, repair, verify, compare, or publish a baseline and its metrics.
| name | drclaw |
| description | Dr. Claw workspace skill for project lookup, session inspection, TaskMaster progress, OpenClaw structured schema, and event-driven reporting |
Use this skill when the user asks about Dr. Claw projects, wants to inspect Claude/Cursor/Codex/Gemini sessions, needs task progress pushed to OpenClaw/mobile, or wants structured OpenClaw-ready JSON outputs.
Before using Dr. Claw, verify the server is reachable:
drclaw server status
If needed, start it:
drclaw server on
drclaw --json projects list
Project references accepted by the CLI:
namedisplayNamepath / fullPathIf a path exists locally but is not registered yet:
drclaw projects add /absolute/path/to/project --name "Display Name"
List sessions:
drclaw --json sessions list <project-ref>
drclaw --json sessions list <project-ref> --provider cursor
Fetch messages:
drclaw --json sessions messages <project-ref> <session-id> --provider claude --limit 100
Send Claude a message:
drclaw --json chat send --project <project-ref> --message "<user message>"
Reply to an existing session with structured OpenClaw output:
drclaw --json chat reply --project <project-ref> --session <session-id> -m "<user message>"
List active sessions across projects:
drclaw --json chat sessions
Check whether TaskMaster is present:
drclaw --json taskmaster detect <project-ref>
Get progress and next action:
drclaw --json taskmaster summary <project-ref>
drclaw --json taskmaster next-guidance <project-ref>
Initialize .pipeline for a project if needed:
drclaw taskmaster init <project-ref>
Configure the default push channel once:
drclaw openclaw configure --push-channel feishu:<chat_id>
Preview a mobile report:
drclaw --json openclaw report --project <project-ref> --dry-run
Send it:
drclaw openclaw report --project <project-ref>
Start the event-driven watcher daemon:
drclaw --json openclaw-watch on --to feishu:<chat_id>
drclaw --json openclaw-watch status
drclaw --json openclaw-watch off
The watcher is now a useful notification pipeline rather than raw websocket forwarding. It:
--deliverCurrent attention-worthy signals include:
human_decision_neededwaiting_for_humanblocker_detectedblocker_clearedtask_completednext_task_changedattention_neededsession_abortedWatcher runtime files:
~/.drclaw/openclaw-watcher-state.json~/.drclaw/logs/openclaw-watcher.logMajor JSON commands now include a top-level openclaw field with a stable versioned schema for mobile / voice clients.
Current schema families:
openclaw.turn.v1openclaw.project.v1openclaw.portfolio.v1openclaw.daily.v1openclaw.report.v1openclaw.event.v1Practical client rules:
decision.needed over guessing whether to interrupt the usernext_actions for quick actions and voice suggestionsturn.summary or portfolio focus for compact renderingopenclaw.event.v1.event.signals first instead of raw typeFormal contract:
cat agent-harness/cli_anything/drclaw/SCHEMA.md
projects list and resolve the project first.chat send or chat reply, and prefer the openclaw schema field over parsing raw reply text.workflow status, digest project, digest portfolio, and taskmaster next-guidance.openclaw report.openclaw-watch on instead of polling digest commands manually.