Harness overview
Harnesses are the reason FlightPath can keep one workflow model while talking to very different AI assistants.
At a glance
| Harness | Runs where | Best for | Notes |
|---|---|---|---|
copilot-cli | CLI | Default path for GitHub users | Good general-purpose default |
claude-cli | CLI | Long-form reasoning and coding | Strong session continuity |
codex-cli | CLI | OpenAI Codex workflows | Exposes Codex-specific sandbox controls |
opencode-cli | CLI | Provider-flexible OpenCode setups | Useful when OpenCode is already your standard |
qwen-cli | CLI | Qwen Code users | Similar workflow model to other CLI harnesses |
gemini-cli | CLI | Gemini-based coding workflows | Supports MCP config support |
copilot-chat | VS Code chat UI | Users who want to stay in the chat window | No standalone CLI equivalent |
Choosing a starting harness
- Start with
copilot-cliif you want the path FlightPath is built around by default. - Use
claude-cliif your team already works in Claude Code. - Use
copilot-chatif you want a VS Code-native experience and do not need the standalone CLI. - Reach for the other CLI harnesses when they already match your preferred assistant or hosting model.
Capability differences
All harnesses use the same workflow schema, but they do not expose identical runtime behavior.
| Capability | CLI harnesses | copilot-chat |
|---|---|---|
| Interactive terminal mode | Yes | No |
| Headless execution | Yes | No |
| Per-harness CLI args | Yes | No |
| Chat UI integration | No | Yes |
| CLI session IDs | Yes | No |
See the per-harness pages for setup notes and tradeoffs.