forensics

AI Writing Forensics Pipeline

Usage

forensics [OPTIONS] COMMAND [ARGS]...

Option	Description
`-V, --version`	Show version and exit
`-v, --verbose`	Enable DEBUG logging
`--no-progress`	Disable Rich progress bars and pipeline observers (CI, logs, minimal TTY).
`--output [text\|json]`	text (default): human-readable lines on stdout. json: one JSON envelope object on stdout (logs to stderr). [default: text]
`--non-interactive`	Disable TUI fallbacks and refuse any prompt; auto-fail if a prompt would block.
`-y, --yes`	Bypass any confirmation prompt.
`--log-format [text\|json]`	Log output format: plain text (default) or JSON for aggregation. [default: text]
`--install-completion`	Install completion for the current shell.
`--show-completion`	Show completion for the current shell, to copy it or customize the installation.
`--help`	Show this message and exit.

Command	Description
`all`	Run full pipeline end-to-end: scrape → extract → analyze → report.
`analyze`	Run analysis pipeline (change-point, drift, convergence, comparison).
`calibrate`	Validate detection accuracy against synthetic ground truth.
`commands`	Dump the full command catalog (for agent discovery).
`config`	Inspect and validate configuration.
`dashboard`	Full-screen pipeline dashboard (`tui` extra). Incompatible with `--no-progress`.
`dedup`	Near-duplicate fingerprint utilities
`export`	Export SQLite + features + analysis into a single `.duckdb` file.
`extract`	Run feature extraction pipeline.
`features`	Feature-store maintenance commands (schema migrations, etc).
`lock-preregistration`	Lock analysis thresholds for pre-registration (run before analyzing data).
`migrate`	Apply pending SQLite migrations (Phase 15 Step 0.2).
`peer-setup`	Print peer setup hints (uv extras, spaCy, Quarto, Ollama pulls from config).
`preflight`	Run preflight checks and report pass/warn/fail for each boundary.
`report`	Render Quarto forensic book.
`scrape`	Crawl and fetch articles for configured authors
`setup`	Launch the interactive setup wizard (requires the ‘tui’ extra).
`survey`	Blind newsroom-wide AI-adoption survey.
`validate`	Validate config.toml, run preflight, and optionally probe live endpoints.