Yuli tools · Yuli Code

The catalog of yuli_* tools your agent receives, by family and with usage examples.

The yuli_* tools are how an external agent uses the IDE. In Claude Code they appear as mcp__yuli__yuli_*. They all go through the same guards as Yuli's own agent, and their results show up in the IDE's panels.

You don't have to name them. Ask for the result (“build and test”, “who calls this method?”) and the agent picks the tool.

Build and quality

Tool	What it does
`yuli_build`	Build, run, test or package with the project's build tool.
`yuli_run_tests`	Runs the tests; the result appears in the Tests panel.
`yuli_verify`	Full verification: build, all tests and one start with the default profile.
`yuli_health`	What's failing: problems, run output, tests, terminals and logs.
`yuli_stop`	Stops whatever is running.
`yuli_diagnostics`	Project errors and warnings.
`yuli_quality_metrics`	Measurement against N−1: `baseline`, `measure`, `after`.
`yuli_review_changes`	Review of the current diff.
`yuli_security_scan` / `yuli_security_center`	Security analysis.
`yuli_governance` / `yuli_governance_write`	Contract, gate and governance documents.

Example: “Verify that everything compiles and the tests pass” → yuli_verify.

Code

Tool	What it does
`yuli_outline`	Structure of a file.
`yuli_search_symbols` / `yuli_get_symbol`	Finds classes, methods and fields.
`yuli_find_references`	Every usage of a symbol.
`yuli_hover`	Type and documentation at a position.
`yuli_impact`	What's affected if you change a file.
`yuli_repo_map` / `yuli_code_search`	Repository map and search.
`yuli_refactor`	IDE refactorings (rename, extract, move).
`yuli_check_structure` / `yuli_check_naming`	Checks the project structure and naming.
`yuli_read_document`	Reads a specification in PDF or Word.

Example: “Rename UserDao to UserRepository across the whole project” → yuli_refactor.

IDE panels

Tool	What it does
`yuli_open_file` / `yuli_show`	Opens a file or a panel in Yuli.
`yuli_terminal_output` / `yuli_run_command`	Reads the terminal or runs a command (with approval).
`yuli_base_url`	The URL where the application started.
`yuli_rest_*`	Creates, lists and sends REST Client requests.
`yuli_diagram`	Generates system diagrams.
`yuli_db_schema` / `yuli_db_query`	Schema and read-only queries.
`yuli_db_execute`	Statements that write, with your approval. See Secure database.

Git

yuli_git_status, yuli_git_commit and yuli_git_push. Commit and push only work if you asked for them, and each one requests approval.

Mission and memory

Tool	What it does
`yuli_mission_get` / `yuli_mission_update`	Reads and advances the mission and its phase.
`yuli_progress`	Mirrors the agent's plan on Yuli's workboard.
`yuli_memory` / `yuli_remember` / `yuli_forget`	Project memory.
`yuli_report`	Sends you a delivery, a question or a progress update.
`yuli_status`	Status of the connection with Yuli.

Web Studio

Family	What it does
`yuli_front_*`	Components, routes, styles, API calls, impact and accessibility.
`yuli_design_*`	Designs, versions, design system and code from the design.
`yuli_preview_*`	Preview without running, screenshots, inspection and API simulator.
`yuli_compound_*`	Brings up the complete full-stack application.

Example: “Design the login screen and bring it to React” → yuli_design_propose, then yuli_design_generate_code.

Agent team

yuli_team_plan, yuli_team_start, yuli_team_status and similar tools plan, launch and track the agent team. The team works with Yuli's engine and the provider configured in Platform mode, not with your subscription.

The full IDE

Family	What it does
`yuli_ide_*`	Manual, commands, settings, shortcuts, open project, configure models and MCP.
`yuli_settings_get` / `yuli_settings_set`	Reads and changes settings.
`yuli_env_*`	Checks and defines environment variables; asks the person for the secret ones.
`yuli_run_config_*` / `yuli_run_launch`	Run configurations.
`yuli_remote_*`	Remote development: open, read, write, run and transfer.

Documents

yuli_export_document (docx, pdf, xlsx), yuli_word_document_outline, yuli_edit_word_document (with preview and approval) and yuli_read_visual, which hands the actual image to the agent.

Connectors

Each tool from your MCP servers appears as yuli_mcp_<service>_<tool>, for example yuli_mcp_github_create_issue. There's also yuli_cloudbees for CloudBees CD/RO. Read tools run freely; tools that write ask for approval.

What it doesn't receive

Six internal tools from Yuli's own cycle: proposing changes, launching subagents, searching for tools, asking questions, publishing the plan and updating it. Your agent uses its own.