Skip to content

COMMANDS

github-actions[bot] edited this page Sep 26, 2026 · 1 revision

Command reference

Run commands from the repository root. Except for the installer, use the managed interpreter exactly as shown:

.\.venv\Scripts\python.exe -m logicytics [global options] <command> [command options]

An option may appear before or after a command when the command supports it. Repeated options may be supplied more than once. plan is the safe checkpoint: it resolves scope without collecting evidence.

Global options

Flag What it does
--config PATH Selects the authoritative YAML configuration. The path is validated before planning.
--usb [DRIVE] Targets a Windows installation on removable storage; omit DRIVE to scan, or provide a drive letter.
--usage Displays local interaction statistics and writes the usage graph.
--modes Prints and writes the generated execution-mode inclusion matrix.
--match TEXT Suggests the closest documented action from a natural-language description.
-h, --help Shows parser help without collecting evidence.

Installer

python -m logicytics.cli.installer [--environment PATH] [--overwrite-config]

The installer is the only entry point intended for a system Python. It creates or reuses the virtual environment and creates logicytics.yaml if absent.

Flag What it does
--environment PATH Uses this virtual-environment directory instead of .venv; a relative path is resolved from the repository root.
--overwrite-config Replaces the root configuration template. Back up a customized file first.

preflight

.\.venv\Scripts\python.exe -m logicytics preflight [flags]

Validates core and, when enabled, plugin sources before collection. It reports valid collectors, quarantined invalid optional plugins, and blocking selected failures.

Flag What it does
--config PATH, --usb [DRIVE] Applies the global configuration or removable-media source selection.
--profile MODE Uses one named mode for selected-plugin validation context.
--include ID Repeats to select exact collector IDs; selected invalid plugins become blocking failures.
--exclude ID Repeats to remove IDs from the request.
--plugins Enables valid PluginCollector implementations from plugins/ for the request.
--invalidate-cache Discards cached validation metadata and performs a fresh source/runtime probe.
--workers COUNT Supplies the bounded worker count carried into request validation.
--block-capability CAP Repeats to prohibit a declared capability. See Safety.

plan

.\.venv\Scripts\python.exe -m logicytics plan --mode standard [flags]

Builds and prints a deterministic, dependency-ordered plan. It starts no collector worker and writes no evidence artifacts.

Flag What it does
--mode {quick,balanced,standard,offline,thorough} Selects the request mode and its scheduling behavior.
--profile MODE Alternative named mode selection. Do not combine it with --mode.
--include ID / --exclude ID Repeats to refine the resolved profile with exact dotted collector IDs.
--plugins Allows valid opt-in plugins to be considered; plugins remain opt-in even when their profile matches.
--invalidate-cache Revalidates collector metadata before planning.
--workers COUNT Sets a positive bounded worker count for the resulting request.
--block-capability CAP Rejects a plan containing a collector that declares the capability.
--config PATH, --usb [DRIVE] Selects configuration or removable-media context.

run

.\.venv\Scripts\python.exe -m logicytics run --mode standard --acknowledge-authorization [flags]

Runs the reviewed plan in isolated workers. Read the final manifest even after an exit code of 0; a package is not proof that every collector succeeded.

Flag What it does
--mode MODE Selects a named execution mode.
--default, --threaded, --minimal, --depth Historical aliases for standard, balanced, quick, and thorough; exactly one mode selector is allowed.
--profile MODE Alternative named mode selector; cannot be combined with a mode selector.
--include ID / --exclude ID Repeats to add or remove exact collector IDs.
--plugins Enables valid opt-in plugins for this run.
--workers COUNT Caps concurrent isolated workers. Sequential work requires 1; parallel work requires at least 2.
--sequential / --parallel Mutually exclusive scheduling override. Use sequential execution for deterministic diagnosis.
--performance-check Forces serial measurement and writes per-collector duration evidence. It cannot be combined with --parallel.
--block-capability CAP Repeats to prohibit declared access categories.
--rerun-from PATH Re-runs explicitly included IDs from a finalized compatible manifest or run directory.
--no-package Finalizes the manifest-backed run directory without writing a ZIP package.
--reboot / --shutdown Mutually exclusive post-run host actions, scheduled only after verified package publication.
--acknowledge-authorization Required confirmation that the operator has authority for the reviewed scope.
--interactive Keeps the interactive command window open after final status.
--config PATH, --usb [DRIVE] Selects configuration or removable-media context.

collector

.\.venv\Scripts\python.exe -m logicytics collector core.system.system_info --acknowledge-authorization [flags]

Runs one exact collector ID independently. It is the preferred way to diagnose one source after reviewing its capabilities.

Flag What it does
COLLECTOR_ID Required positional dotted ID for the one collector to run.
--acknowledge-authorization Required authorization confirmation.
--no-package Keeps the manifest-backed result without a ZIP.
--interactive Keeps an interactive command window open at completion.
--workers COUNT Sets the bounded worker limit; the direct collector still resolves one selected source.
--block-capability CAP Rejects the collector when it declares a blocked capability.
--config PATH, --usb [DRIVE] Selects configuration or removable-media context.

collector cannot be combined with --profile, --include, --exclude, or --plugins; its positional ID is the complete selection.

debug

.\.venv\Scripts\python.exe -m logicytics debug [flags]

Writes a bounded diagnostic JSON report containing environment, interpreter, configuration, maintenance, optional-tool, and preflight information. Sanitize it before sharing.

Flag What it does
--config PATH, --usb [DRIVE] Uses the selected configuration and target context while gathering diagnostics.
--profile MODE, --include ID, --exclude ID, --plugins Applies the request selection used for preflight diagnostics.
--workers COUNT, --block-capability CAP Carries execution and policy context into the diagnostic request.

update

.\.venv\Scripts\python.exe -m logicytics update [--apply] [--launch-action ACTION --new-window]

Checks Git, the repository, origin configuration, and remote reachability. It never pulls unless --apply is explicit.

Flag What it does
--apply Runs git pull after successful repository and remote checks.
--launch-action {preflight,debug,dev} Selects an allowlisted maintenance action to launch after update. Requires --new-window.
--new-window Starts the allowlisted action in a visible separate Windows console.
--config PATH, --usb [DRIVE] Accepts common context flags; update behavior itself operates on the checkout.

dev

.\.venv\Scripts\python.exe -m logicytics dev [--write-manifest --next-version VERSION] [--interactive]

Runs repository integrity and organization checks. It is a developer maintenance command, not an evidence collection command.

Flag What it does
--write-manifest Writes the reviewed local integrity manifest; requires --next-version unless interactive mode supplies it.
--next-version VERSION Provides the semantic version recorded in a newly written manifest.
--interactive Shows checks and prompts before an integrity-manifest write.
--config PATH, --usb [DRIVE] Accepts common context flags.

Exit codes

Code Meaning Next action
0 Command completed successfully. Review the manifest or generated report.
1 A collection finished with non-success status. Preserve the manifest and inspect partial, skipped, failed, or cancelled records.
2 Argument, configuration, preflight, planning, or command error. Read the rendered error and Error reference; correct the cause rather than bypassing validation.
130 Interrupted. Locate the durable manifest and determine what safely finalized before retrying.

Clone this wiki locally