Flightops is in beta (0.2.0b1) for invited ROS 2 teams. Read the beta first, then install and follow first steps. Use the user guide for everyday workflows with expected results and troubleshooting. Use the task guides below for a specific result.
The beta
| Page | Use it for |
|---|---|
| The beta | What is included, known limits, feedback |
| Release notes | What changed in 0.2.0b1 |
| Privacy notice | What the beta service keeps and for how long |
| License and usage metrics | Activation, renewal, devices and the metrics that are sent |
Tutorials
| Task | Guide |
|---|---|
| Open Flightops and learn the controls | First steps |
| Test on fresh Ubuntu with your own package and TurtleBot3 | Fresh Ubuntu demo |
| Attach to ROS nodes that are already running | Attach |
| Start and stop a ROS launch file | Launch |
| Diagnose real Nav2, Panda, and Go2 simulation faults | ROS 2 troubleshooting examples (offline HTML) |
| Set up optional AI | AI setup |
| Open and keep a results report | Results |
| Read a recording without ROS | Recording quickstart |
| Record a demo topic, query it, and reopen its report | Capture to report |
| Read the terminal and report screens | Screen tour |
| Replay a recording or compare two runs in the desktop app | Scene Workbench |
| Run a test session | Testing checklist |
AI assistance
- AI overview.
- Set up a provider.
- Use the assistant.
- Connections, credentials, and troubleshooting.
- Optional Jev evidence priority.
- Launch log summary and recall: read and search a launch log in the terminal.
Task guides
| Task | Guide |
|---|---|
| Check topic delivery or lifecycle state after launch | Launch readiness |
| Trace a map, lidar, planner, joint, or control-topic fault | ROS 2 troubleshooting examples |
| Investigate a ROS project from source to live observations | Development workspace |
| Compare parameters and inspect message-time transforms | Development tools |
| Plan and retain a development experiment | Development experiments |
| Save and reopen project settings | Project profiles |
| Resolve a selected launch file and its declarations | Launch resolution |
| Set up the launch gate for a project | Project intake |
| Inspect parameter provenance, timelines, or change impact | Investigation workflows |
| Inspect laser, IMU, or depth inputs | Device recipes |
| Run repeated seeded simulation experiments | Simulation experiments |
| Test a depth-camera pipeline | Depth camera |
| Run an agent development experiment | Developer test lab |
| Compare simulation runs or inject a bounded fault | Behavior and fault workflows |
| Prepare a recording for replay | Replay readiness |
| Compare two saved proof records | Proof comparison |
| Package evidence for another machine | Reproduction packages |
| Deliver to a remote development container | Remote development |
Reference
| Reference | Use it for |
|---|---|
| Command line | Public commands and common options |
| Build features | What the beta build includes |
| License and usage metrics | Activation, renewal, devices and metrics |
| Glossary | Plain-language definitions and term links |
| Terminal workspace | Views and keyboard controls |
| Integrations | MCP setup, results reports, and source builds |
| Scene Workbench protocol | The local service protocol behind the Scene Workbench |
| Bounded evidence queries | MCAP query clocks, limits, and result meaning |
| TF history | Session TF collection, clocks, and bounds |
| Troubleshooting | Common setup and launch problems |
| Supported ROS and Gazebo environments | Supported development images |
| Scenario format | Versioned scenario fields and semantics |
| Security boundaries | Execution, evidence, and hardware limits |
Reference pages by topic
Each page describes one command family or data format. The command reference maps every flightops command to its page.
| Topic | Page | Use it for |
|---|---|---|
| Recording | Managed recording | How a launch arms bounded capture, and its limits |
| Recording | Managed recording commands | flightops record enable, list, show, and view |
| Recording | Independent collector | A separate exact-topic recorder that does not control the application |
| Recording | Retention | Scoped budgets, pins, read leases, and expiry plans |
| Recording | Recording reduction | Shrinking one recorded failure to a bounded replay |
| Recording | Evidence resolver | Digest-bound receipts and bounded recording segments |
| History | History commands | flightops history and flightops checks |
| History | History catalog | The local SQLite index and its record shapes |
| History | Knowledge | Cases and immutable strategy revisions |
| History | Owned sources | Imported owned text files as inert sources |
| History | Governed learning | Evaluation suites and conditional training decisions |
| History | Retrospective AI | Reviewed disclosure and one optional provider turn |
| Checks | Check catalog | The three bounded sensor observations |
| Checks | Check ranking | Deterministic ordering and run-check |
| Checks | Change review | Comparing snapshots to select affected checks |
| Checks | Integration contracts | Frozen requirements evaluated against saved observations |
| Workspace and sharing | Personal workspace | The private loopback desktop workspace |
| Workspace and sharing | Structured handoff | Moving a task timeline between users in quarantine |
Component tools
| Tool | Guide |
|---|---|
| Live graph and topic inspection | onscope |
| Recording analysis and replay fixtures | splashdown |
| Build, test, and simulation gates | launchcommit |
| Reusable build workspaces | highbay |
| ROS skills for coding agents | flightschool |
| Coordinator directory for coding agents (preview) | flightlead |
| MCP server and results reports | integrations |
| Desktop app for recordings | Scene Workbench |
| Local KVM smoke task (research preview) | Gym |
| Shared data and evidence library | axikit |
| Recording-to-regression workflow | Recorded failures |
Project guides
- Demo project explains the bundled ROS example.
- Safety model holds the safety requirements that outrank every other design choice.
- Known limits lists what the beta does not do and what has not been checked.
For a short overview of the component workflows, use the advanced guide.