Trae SOLO

Control Trae SOLO from OpenCLI through the Electron debug port and read its local VSCode-style state files.

Mode: Desktop app / local filesystem · App: TRAE SOLO

Commands

Command	Description	Access
opencli trae-solo status	Check whether the Trae SOLO desktop app is reachable	read
opencli trae-solo history	List visible projects and tasks from the Trae SOLO sidebar	read
opencli trae-solo model [name]	Read, list, or switch the active model in an open task	write when switching
opencli trae-solo mode [code|work]	Read or switch between Code and Work mode	write when switching
opencli trae-solo automation-list	Read visible Automation tab entries	read
opencli trae-solo skill-list	List marketplace or installed skills from the Skills panel	read
opencli trae-solo skill-search <keyword>	Search visible marketplace skills	read
opencli trae-solo skill-category [name]	List or filter marketplace skill categories	read
opencli trae-solo storage-keys	List renderer localStorage / sessionStorage keys	read
opencli trae-solo storage-get <key>	Read a renderer storage value	read
opencli trae-solo cookies	List JavaScript-visible renderer cookies with truncated previews	read
opencli trae-solo idb-list	List renderer IndexedDB database names	read
opencli trae-solo state-keys	List keys in Trae SOLO state.vscdb	read
opencli trae-solo state-get <key>	Read a key from state.vscdb	read
opencli trae-solo recent-workspaces	Show recently opened workspaces from local state	read
opencli trae-solo workspaces-list	List workspaceStorage entries and resolved workspace targets	read
opencli trae-solo extensions-list	List installed VSCode-compatible extensions	read
opencli trae-solo task-fs-list	List on-disk Trae SOLO task ids	read
opencli trae-solo task-fs-turns <task-id>	List chat-turn git tags for a task snapshot	read
opencli trae-solo task-fs-show <task-id>	Show the workspace tree at a chat-turn ref	read
opencli trae-solo skill-fs-list	List local skill directories under ~/.trae/skills	read
opencli trae-solo skill-fs-installed	List skills registered in skill-config.json	read
opencli trae-solo skill-fs-show <name>	Show a local skill's SKILL.md head and metadata	read
opencli trae-solo settings-read	Read user settings.json with JSONC comments/trailing commas	read
opencli trae-solo user-rules	Read ~/.trae/user_rules.md	read

Write-side UI commands that only proved button clicks were intentionally left out. New task creation, task open/navigation, message actions, skill install/uninstall/run/toggle, automation creation, and filesystem deletion need explicit postconditions before they can be exposed safely.

Examples

opencli trae-solo status
opencli trae-solo history --limit 20
opencli trae-solo model --list true
opencli trae-solo model "Claude"
opencli trae-solo mode work

opencli trae-solo skill-search "python"
opencli trae-solo automation-list --tab task-template

opencli trae-solo state-keys --filter workbench
opencli trae-solo recent-workspaces
opencli trae-solo task-fs-list --limit 20

Notes

• Electron UI commands require Trae SOLO to be running with the configured CDP port. OpenCLI launches registered Electron apps with the app-specific debug port when needed.
• Renderer storage reads come from the current Electron renderer and may be empty if the app has not loaded the relevant workspace.
• Filesystem reads are local-only and read Trae SOLO state under ~/.trae and ~/Library/Application Support/TRAE SOLO.
• model and mode verify the visible post-action state before returning success.