CLI Reference¶
The fcc command-line tool provides commands for project scaffolding, validation, simulation, and documentation generation. It is built with Click and installed as an entry point via pip install -e ..
Global Options¶
Commands¶
fcc init¶
Initialize a new FCC project with scaffolded directory structure.
Synopsis:
Options:
| Option | Required | Default | Description |
|---|---|---|---|
--name |
Yes | -- | Project name |
--dir |
No | . |
Target directory for the new project |
What it creates:
fcc_config.yaml-- Project configurationpyproject.toml-- Python project filesrc/-- Source code directorytests/-- Test directorydata/-- Data files directory (personas, workflows, schemas, scenarios, governance)Makefile-- Build automation
Example:
fcc init --name my-doc-project --dir /path/to/workspace
cd /path/to/workspace
make venv && source .venv/bin/activate && make install-dev
fcc add-persona¶
Add a new persona to an existing FCC project.
Synopsis:
Arguments:
| Argument | Description |
|---|---|
name |
Full persona name (e.g., "Custom Analyst") |
Options:
| Option | Required | Default | Description |
|---|---|---|---|
--phase |
Yes | -- | FCC phase: Find, Create, or Critique |
--id |
Yes | -- | Short identifier (e.g., CA) |
--dir |
No | . |
Project directory |
Example:
fcc validate¶
Validate the FCC project structure, checking for required files and directories.
Synopsis:
Options:
| Option | Required | Default | Description |
|---|---|---|---|
--dir |
No | . |
Project directory to validate |
Checks performed:
fcc_config.yamlexistspyproject.tomlexistssrc/directory existstests/directory existsdata/directory exists
Exit codes:
0-- Validation passed1-- One or more validation issues found
Example:
fcc simulate¶
Run an FCC workflow simulation using either mock (deterministic) or AI-powered mode.
Synopsis:
Options:
| Option | Required | Default | Description |
|---|---|---|---|
--scenario |
No | GEN-001 |
Scenario ID to run |
--mock / --no-mock |
No | --mock |
Use mock AI client (deterministic) or real AI |
--dir |
No | . |
Project directory |
How it works:
- Loads the workflow graph from
data/workflows/base_sequence.json. - Creates an
AIClientwith either the mock provider or a real provider (Anthropic/OpenAI). - Constructs an
AISimulationEnginewithmax_steps=20. - Runs the simulation starting from the
RCnode. - Writes traces to
traces_ai.json.
Example:
# Run with mock client (default, no API key needed)
fcc simulate --scenario GEN-001
# Run with real AI client
fcc simulate --no-mock --scenario ADV-001
fcc generate-docs¶
Generate docs-as-code documentation from persona specifications using Jinja2 templates.
Synopsis:
Options:
| Option | Required | Default | Description |
|---|---|---|---|
--dir |
No | docs |
Output directory for generated documentation |
--personas |
No | all |
Filter: all, core, integration, governance, stakeholder, champions, or a single persona ID |
Output:
Generates 56 documentation files per persona (tutorials, prompts, workflows) across the filtered set. With all 24 personas, this produces 1,348 files.
Example:
# Generate docs for all personas
fcc generate-docs --dir docs/generated
# Generate docs for core personas only
fcc generate-docs --personas core
# Generate docs for a single persona
fcc generate-docs --personas RC
fcc validate-docs¶
Validate generated documentation for completeness, checking for empty files.
Synopsis:
Options:
| Option | Required | Default | Description |
|---|---|---|---|
--dir |
No | docs |
Documentation directory to validate |
Checks performed:
- Scans
<dir>/fcc/personas/recursively for.mdfiles. - Flags any empty files.
Example:
fcc sitemap¶
Generate a SITEMAP.md file for the documentation tree.
Synopsis:
Options:
| Option | Required | Default | Description |
|---|---|---|---|
--dir |
No | docs |
Output directory |
Output:
Generates <dir>/fcc/SITEMAP.md with a linked index of all generated documentation files.
Example:
Environment Variables¶
| Variable | Purpose |
|---|---|
ANTHROPIC_API_KEY |
Required for --no-mock simulation with Anthropic |
OPENAI_API_KEY |
Required for --no-mock simulation with OpenAI |
Related Pages¶
- Architecture -- the modules behind each command
- Extension Guide -- adding custom commands
- Testing Guide -- testing CLI commands
- Workflow Graphs -- the graphs that
simulatetraverses