rosclaw ยท diff

git:20260721.d4f583d to git:20260722.ef835bc

85 added, 20 removed. Audit A to A.

---
name: rosclaw
description: Use when operating, validating, or changing ROSClaw physical-AI runtime workflows, especially CLI smoke tests, Practice evidence loops, body/runtime checks, MCP integration, MuJoCo sandbox verification, and safe ROS or hardware boundaries.
---
# ROSClaw Agent Skill
- Use this exact launcher for every ROSClaw CLI command: `rosclaw`. Do not replace
+ Use this exact launcher for every ROSClaw CLI command: `/code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint`. Do not replace
it with a different `rosclaw` found on `PATH`.
## Safety
- Treat ROSClaw as physical-AI infrastructure. Do not publish ROS topics,
actuate hardware, run real robot skills, or mutate a live workspace directly,
even when a user asks for that specific action.
- Prefer dry-run, read-only, mock, fixture, simulation, or temp-workspace
commands for validation.
- Use a temporary `ROSCLAW_HOME` for CLI smoke tests that write persistent state.
- Prefer `--json` for machine checks, then validate the JSON parses.
- The only Agent-side action entry is MCP `request_action`, backed by
`rosclawd`. Never instantiate a local Runtime or register a hardware
driver/executor from the Agent process.
- REAL actions require an immutable body snapshot, a daemon-issued
capability- and action-intent-bound permit, and a registered verified REAL
executor. Refuse direct ROS, DDS, serial, CAN, SDK, or motor commands.
## First Checks
```bash
- rosclaw doctor --json
- rosclaw status capabilities
- rosclaw daemon status --json
- rosclaw demo list
- rosclaw demo run ur5e-reach
- rosclaw explain latest
- rosclaw agent doctor universal --project-root .
- rosclaw agent test universal --project-root . --quick --mcp-probe
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint doctor --json
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint status capabilities
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint daemon status --json
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint demo list
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint demo run ur5e-reach
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint explain latest
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint agent doctor universal --project-root .
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint agent test universal --project-root . --quick --mcp-probe
```
For `daemon status`, require `running` and `ledger.integrity_verified` to be
`true`; require `ledger.write_failed`, `recovery.required`, and
`emergency_stop_latched` to be `false`. Production REAL work also requires
`supervision_state=ARMED`, `privilege_separated=true`, and
- `rosclaw daemon security-check --json` with
+ `/code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint daemon security-check --json` with
`boundary_ready=true`, `daemon_uid_pinned=true`, and
`ledger_state_private=true`; same-UID development proves only process separation.
- Read an existing durable result with `rosclaw daemon action-status <ACTION_ID>
- --json` and `rosclaw daemon receipt <ACTION_ID> --json`. Do not use
+ Read an existing durable result with `/code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint daemon action-status <ACTION_ID>
+ --json` and `/code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint daemon receipt <ACTION_ID> --json`. Do not use
`daemon acknowledge-recovery` as Agent automation: it is an operator
incident-review command and does not clear E-stop or prove physical state.
## Robot Integration Workflow
Robot Integration installation/configuration is a CLI lifecycle around the MCP
runtime. It is not an additional MCP tool surface.
```bash
- rosclaw robot discover --json
- rosclaw robot install realsense --json
- rosclaw robot verify realsense --stage contract --json
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint robot discover --json
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint robot install realsense --json
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint robot verify realsense --stage contract --json
```
- `robot install` verifies and installs the signed contract. Pass
`--install-adapter` only when the operator explicitly authorizes dependency
installation in the selected Python environment.
- Bind a discovered device with `robot configure` only when an exact model and
stable serial are available. `--allow-offline` creates configuration, never
hardware evidence.
- A read-only hardware check requires a configured instance and canonical
`rosclawd` receipt. Treat local success as candidate evidence; never promote
product support or claim physical success from CLI text alone.
- Do not call the vendor SDK directly. Runtime device access stays behind MCP
`request_action` and the daemon boundary.
## Capability App Workflow
```bash
- rosclaw app list
- rosclaw app validate realsense-inspect --json
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint app list
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint app validate realsense-inspect --json
```
- Apps call named Capabilities only. They do not install drivers, issue
permits, arm rosclawd, or authorize direct hardware access.
- Run SHADOW first with `app run <APP> --body <BODY> --mode SHADOW --json`.
Treat a local success as component evidence, never an independent hardware
or Agent verification claim.
## Practice Evidence Loop
Use fixture-based Practice workflows before touching real robots:
```bash
TMP=$(mktemp -d /tmp/rosclaw-practice.XXXXXX)
export ROSCLAW_HOME="$TMP/home"
- rosclaw practice record --fixture tests/fixtures/practice/rh56_minimal_loop.json --out "$TMP/practice" --json
- rosclaw practice verify practice_rh56_minimal_loop --data-root "$TMP/practice" --strict --json
- rosclaw practice distill practice_rh56_minimal_loop --data-root "$TMP/practice" --json
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint practice record --fixture tests/fixtures/practice/rh56_minimal_loop.json --out "$TMP/practice" --json
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint practice verify practice_rh56_minimal_loop --data-root "$TMP/practice" --strict --json
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint practice distill practice_rh56_minimal_loop --data-root "$TMP/practice" --json
```
## MCP Contract
- Expected P0 tools: `get_robot_state`, `list_skills`, `query_memory`, `validate_trajectory`, `sandbox_run`, `practice_query`, `emergency_stop`, `get_body_profile`, `get_body_state`, `list_body_capabilities`, `query_body`, `validate_body_action`, `get_calibration_status`, `get_runtime_status`, `request_action`, `get_action_status`, `cancel_action`, `get_product_status`, `list_product_demos`, `run_product_demo`, `get_execution_receipt`, `explain_execution`.
- `run_product_demo` executes an official MuJoCo path and persists an
integrity-checked `ExecutionReceipt`; it never commands real hardware.
- Use `get_execution_receipt` and `explain_execution` instead of inferring
success from text output.
- `sandbox_run` is simulation-only. If no physics state is available, treat the
result as degraded rather than live.
- `validate_trajectory` can approve a plan, but it does not authorize direct
hardware execution.
- `request_action` sends an unapproved envelope to `rosclawd`; only the daemon
may turn a matching server-issued permit into authorization.
- `cancel_action` never claims that active physical motion stopped. Use
`emergency_stop` and the certified hardware E-stop when motion may be active.
+
+ ## LeRobot Bridge Workflow
+
+ ### Operator setup
+
+ These are operator workflows. Do not install dependencies unless explicitly requested.
+
+ ```bash
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint setup lerobot --reference-policy rh56
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint lerobot doctor --json
+ ```
+
+ ### Read-only Agent discovery
+
+ Use MCP tools:
+
+ 1. `get_product_status`
+ 2. `get_runtime_status`
+ 3. `get_body_profile`
+ 4. `get_body_state`
+ 5. `get_calibration_status`
+ 6. `list_body_capabilities`
+
+ ### SHADOW request
+
+ Submit through `request_action` with execution mode SHADOW.
+ Poll with `get_action_status`.
+ Read the final result with `get_execution_receipt`.
+ Explain it with `explain_execution`.
+
+ ### REAL request
+
+ Submit the exact requested action through `request_action`.
+ The Agent does not issue or approve a Permit.
+ If rosclawd reports AUTHORIZATION_REQUIRED, stop and explain the missing authorization.
+ Never fall back to shell, serial, CAN, ROS topic, MCP driver, or vendor SDK execution.
+
+ ### Dataset workflow
+
+ For non-physical data operations, use the ROSClaw CLI:
+
+ ```bash
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint practice verify <PRACTICE_ID> --strict --json
+ /code/rosclaw/rosclaw_lerobot/rosclaw_repo/.venv/bin/python -m rosclaw.entrypoint practice export --practice-id <PRACTICE_ID> --format lerobot --profile physical --output <OUTPUT_DIR> --json
+ ```
+
+ ### Unsupported requests
+
+ Training, DAgger, reward models, Hub publishing, arbitrary-policy robot mapping,
+ CAN RH56 execution, open-loop chunks, and unattended execution are outside
+ LeRobot Bridge v1.0.1.
+
+ ### Decision table
+
+ | User intent | Agent should do | Agent must not do |
+ |---|---|---|
+ | "Is LeRobot installed?" | `get_product_status` + `get_runtime_status` | import lerobot |
+ | "Can I use the RH56 now?" | body/profile/state/calibration/capability tools | open the serial port |
+ | "Preview what it would do" | `request_action(SHADOW)` | run `rollout execute` directly |
+ | "Do the real OK gesture" | `request_action(REAL)`, wait for rosclawd | self-issue a Permit |
+ | "Try it without authorization" | refuse and explain AUTHORIZATION_REQUIRED | switch to serial commands |
+ | "Did it succeed?" | `get_action_status` + receipt + `explain_execution` | guess success from text |
+ | "Export data for training" | `practice verify` / `export` | train automatically |
+ | "Control RH56 with any VLA" | explain v1.0.1 supports only the reference contract | guess mapping by action_dim |
+ | "Control RH56 over CAN" | explain CAN is fail-closed unsupported | reuse RS485 mapping on CAN |