TRAE is a brand of AI development tools combining project context, code changes, commands, and repeatable instructions. It covers TraeCode and TraeWork; check features and terms for each client separately.
Working environment around the task
Changing a working project requires file access, verification commands, and relevant context, rather than just an isolated snippet. TRAE organizes this work around an agent in a development environment. Official materials describe programming tasks, rules, and context; the current brand website offers TraeCode and TraeWork as separate access options.
Materials use the names SOLO, TraeCode, and TraeWork. They refer to modes and components under one brand and do not need duplicate catalog cards. When installing, check the specific client against its documentation: the brand name does not guarantee identical buttons, features, or access terms across variants.
Rules and skills do not replace requirements
Rules suit conventions needed all the time: code style, comment language, and check commands. Skills describe a repeatable workflow and are loaded when needed. The official guide explains the SKILL.md structure and using ready-made instruction packages.
For example, prepare a skill for updating a UI component: locate existing variants, preserve keyboard navigation, check narrow widths, and avoid introducing a new component unnecessarily. Add the specific defect and expected behavior to the prompt. A shared rule set should not conceal page differences or let an agent invent requirements.
Exercise: fix form state
A registration form’s button remains disabled after a network error. Ask TRAE to locate request state handling, trace the failed response, and propose a targeted fix. Require the design and server contract to stay unchanged. After editing, check the error case, successful submission, and retry.
Check keyboard behavior too: focus must remain available and the error message readable. An agent report helps when it identifies changed files and commands run. “Done” alone does not prove that the bug was reproduced or checks passed.
Commands, data boundaries, and terms
TRAE describes sandbox isolation and shell command filtering as controls on agent actions. Their support depends on the client and platform; automatic execution settings need careful review. Permission to bypass the sandbox expands the process’s permissions, so first understand which command needs access and why.
An official pricing page exists, but check changing quotas and full plan details in your account and current documentation. This card does not promise ongoing free use of every model, a Russian interface, or local inference. An installed environment and tool support alone do not establish the product’s own public API. Check model choice and code-sharing rules before working with a private project.
The useful rules of the project are short and verifiable: specify the test command and the expected artifact that the agent must return after it is executed.
Suitable tasks
- Correcting errors in an existing project
- Establish conventions through Rules
- Describe a repeatable check in SKILL.md
- Control commands in agent-driven development
Strengths
- Environment with project context and agent
- Separation of permanent rules and skills
- Documented shell controls
Limitations
- Older mode names do not establish current terms
- Sandbox support depends on the client and platform
- Installation alone does not establish full quotas, local inference, or API access
Selection and conditions
- Free access
- Not confirmed
- Trial period
- Not confirmed
- No sign-up required
- Not confirmed
- Russian interface
- Not confirmed
- Local installation
- Not confirmed
- API
- Not confirmed
“Not confirmed” means there is not enough information. The feature may still be available; check the developer’s terms.
- Input formats
- Text, Code
- Result
- Text, Code
Terms
Changing plans and detailed terms require checking in your account. A local app does not prove local inference; API availability and a Russian interface remain unknown.From task to first result
Practical example
- Task
- Correcting errors in an existing project
- What to prepare
- A small part of your own project and a reproducible description of the error, with no tokens or passwords.
Task brief or prompt
Explain why this function handles an empty list incorrectly. First describe its current behavior, then suggest a minimal fix and checks for empty, normal and invalid input. Do not change other parts of the project unnecessarily.
Expected result
An explanation of the cause, a small proposed diff and checks to run locally.
How to check the result
- Read the diff before applying it.
- Run the checks in a separate working copy and make sure no secrets were included in the prompt.
An editorial teaching example of a task brief, not the result of hands-on testing of the service. Check that the required mode is available, that you have rights to the source material and that the intended use of the result is permitted.
Check details
Official sources
- TRAE: official product
- TRAE: current documentation for TraeCode and TraeWork
- TRAE: a development tutorial using SOLO
- TRAE: Rules and Agent Skills
- TRAE: sandbox and shell command control
- TRAE: current access conditions
This description is based on the developer’s material. Features, prices and availability may change; check the primary source before use.
