Fireteam automatically estimates task complexity to select the appropriate execution strategy.
| Level | Description | Examples |
|---|---|---|
| TRIVIAL | Single-line changes | Typo fixes, adding comments |
| SIMPLE | Self-contained changes | Single-file modifications, simple bug fixes |
| MODERATE | Multi-file changes | Refactoring a module, adding a feature with tests |
| COMPLEX | Architectural changes | Major refactoring, new subsystems |
When you run fireteam start or fireteam run without specifying a mode, Fireteam:
- Sends your goal and context to Claude
- Claude analyzes the scope using read-only tools (Glob, Grep, Read)
- Returns a complexity level based on the analysis
- Fireteam maps the complexity to an execution mode
# Let Fireteam auto-detect complexity
fireteam start
# Force a specific mode
fireteam start -m full| Complexity | Mode | Behavior |
|---|---|---|
| TRIVIAL | SINGLE_TURN | Single CLI call, no review |
| SIMPLE | SINGLE_TURN | Single CLI call, no review |
| MODERATE | MODERATE | Execute + review loop until ≥95% |
| COMPLEX | FULL | Plan + execute + 3 parallel reviews until 2/3 say ≥95% |
- Fix typos
- Add/remove comments
- Rename a single variable
- Simple formatting changes
- Implement a single function
- Add logging to existing code
- Fix a straightforward bug
- Update configuration values
- Refactor a module
- Add a new feature with tests
- Fix a bug requiring investigation
- Update multiple related files
- Major architectural changes
- Implement new subsystems
- Large-scale refactoring
- Cross-cutting concerns
You can bypass complexity estimation by specifying the mode directly:
# Force FULL mode for thorough execution
fireteam start -m full
# Force SINGLE_TURN for quick execution
fireteam start -m single_turnFireteam explores the codebase using read-only tools to make accurate complexity estimates. Providing relevant files in your PROMPT.md helps with this:
# Task
Add logging to the auth module.
## Context
@src/auth/The included files give Claude context for a more accurate complexity assessment.