Choose Your Modules

SpecFact ships as six independent modules. Each solves a distinct problem in your development workflow. You install only what you need.

CLI vs AI IDE: SpecFact CLI is fully deterministic — no AI built in. All commands produce structured, reproducible output. For AI-assisted workflows, each module ships IDE slash prompts that you install with specfact init ide. These prompts feed CLI output to your IDE’s AI (Cursor, Claude Code, VS Code/Copilot, Windsurf, and 10+ more) for interactive analysis and refinement.

Quick Decision Guide

“I need to…“ — find your starting point:

I need to… Install this Command surface
Manage my backlog with template-driven refinement, standups, and sprint planning Backlog specfact backlog
Structure my project, link plans to code, and manage dev lifecycle Project specfact project, specfact plan, specfact sync
Import a legacy codebase, detect features, and track spec drift Codebase specfact code import, specfact code drift
Run governed code reviews with quality scoring and house rules Code Review specfact code review
Validate OpenAPI/AsyncAPI contracts and generate test suites Spec specfact spec validate, specfact spec mock
Enforce quality gates before merge or release Govern specfact govern enforce

Module Overview

Backlog — Template-Driven Backlog Management

The problem it solves: Your team wastes hours in refinement meetings, stories lack acceptance criteria, and nobody tracks what changed between sprints.

What it does:

  • Connects to GitHub Issues or Azure DevOps and syncs your backlog items
  • Runs deterministic refinement: detects story type (user story, defect, spike, enabler), scores completeness confidence, validates Definition of Ready
  • Automates ceremonies: daily standup summaries, refinement checklists, planning assistance, kanban flow tracking
  • Tracks deltas between sprint snapshots: what moved, what’s at risk, estimated cost of changes
  • Enforces backlog policies per framework (Scrum, Kanban, SAFe)
  • Ships IDE slash prompts (via specfact init ide) that feed CLI output to your AI IDE for interactive refinement

Best for: Any team managing a backlog — from solo developers tracking GitHub issues to enterprise teams running SAFe on Azure DevOps.

specfact module install nold-ai/specfact-backlog
specfact backlog --help

Full Backlog docs →


Project — Project Structure and Lifecycle

The problem it solves: Your plans live in documents, your code lives in repos, and nothing connects the two. Feature work has no traceable path from idea to release.

What it does:

  • Creates structured project bundles that link plans to code
  • Manages development plans with feature/story tracking aligned to Software Design Documents (SDDs)
  • Syncs with external tools: OpenSpec, Spec-Kit, GitHub, Linear, Jira
  • Provides a DevOps flow with explicit stages: plan → develop → review → release → monitor
  • Imports brownfield codebases with automatic feature detection

Best for: Teams that want traceable connections between planning artifacts and code, or anyone modernizing a legacy codebase.

specfact module install nold-ai/specfact-project
specfact project --help

Full Project docs →


Codebase — Code Analysis and Drift Detection

The problem it solves: Your implementation drifted from the spec months ago, you have orphaned API endpoints nobody documented, and your test coverage doesn’t match your contracts.

What it does:

  • Imports existing codebases and detects features using AST parsing and semantic analysis
  • Measures OpenAPI contract coverage: which endpoints have specs, which don’t
  • Detects drift between specs and implementation — finds orphaned specs, missing tests, and alignment gaps
  • Validates external codebases via sidecar mode (no source modification)
  • Runs reproducibility suites: lint, type-check, contract validation, and CrossHair property tests

Best for: Teams with existing codebases that need to understand what they have, where specs are missing, and what drifted.

Requires: Project module (installed automatically as dependency).

specfact module install nold-ai/specfact-codebase
specfact code analyze --help
specfact code drift --help

Full Codebase docs →


Code Review — Governed Quality Reviews

The problem it solves: Code reviews are inconsistent, quality is subjective, and there’s no measurable record of improvement over time.

What it does:

  • Runs governed code reviews with deterministic quality scoring (not just linting — structural analysis)
  • Bundles Ruff, Semgrep, Pylint, Pytest, Radon, BasedPyright, and CrossHair for comprehensive analysis
  • Maintains a review ledger: tracks coins/streaks, violations over time, and team statistics
  • Supports house rules per IDE (Cursor, Claude Code, Codex, Windsurf) to enforce consistent quality standards
  • Validates TDD gates: did tests come before implementation?

Best for: Any developer or team that wants measurable, consistent code quality. Works standalone — no other modules required.

specfact module install nold-ai/specfact-code-review
specfact code review run --help

Full Code Review docs →


Spec — Contract Validation and Test Generation

The problem it solves: Your OpenAPI specs exist but nobody validates them against the actual code, breaking changes slip through, and writing contract tests is tedious.

What it does:

  • Validates OpenAPI and AsyncAPI contracts using Specmatic with example verification
  • Detects backward-incompatible changes between spec versions before they reach production
  • Generates test suites directly from your contracts — no manual test writing
  • Spins up mock servers from specs for frontend development and integration testing
  • Bootstraps and validates Software Design Document (SDD) constitutions

Best for: Teams with API contracts that need validation, backward compatibility checks, and generated test suites.

Requires: Project module (installed automatically as dependency).

specfact module install nold-ai/specfact-spec
specfact spec validate --help
specfact spec mock --help

Full Spec docs →


Govern — Quality Gates and Enforcement

The problem it solves: Quality standards exist on paper but nothing enforces them before merge or release. Patches are applied manually and inconsistently.

What it does:

  • Enforces quality gates at configurable strictness levels: minimal, balanced, or strict
  • Validates SDD compliance: checks bundle state, thresholds, and frozen sections before promotion
  • Manages patches with explicit preview → apply → verify flow and write authorization gates
  • Works in both interactive mode and CI/CD pipelines

Best for: Teams that need automated quality enforcement before merges and releases, especially in CI/CD pipelines.

Requires: Project module (installed automatically as dependency).

specfact module install nold-ai/specfact-govern
specfact govern enforce --help

Full Govern docs →


Side-by-Side Comparison

  Backlog Project Codebase Code Review Spec Govern
One-line USP Template-driven backlog refinement & ceremonies Plan-to-code traceability Legacy import & drift detection Governed quality scoring Contract validation & test gen Quality gate enforcement
Dependencies None None Project None Project Project
Works standalone Yes Yes No Yes No No
Connects to external tools GitHub, Azure DevOps GitHub, Linear, Jira, OpenSpec Specmatic
CI/CD ready Yes Yes Yes Yes (score-only mode) Yes Yes (non-interactive)
Solo developer Great Great Great Great Useful Optional
Team use Essential Essential Very useful Essential Very useful Very useful
Enterprise Essential Essential Essential Essential Essential Essential

Solo Developer — Start simple

specfact init --profile solo-developer
# Installs: backlog + code-review

You get template-driven backlog management for your GitHub issues and governed code reviews. Install IDE slash prompts with specfact init ide for AI-assisted workflows. Add more modules as your project grows.

Small Team — Full workflow

specfact init --install backlog,project,code-review,spec

Covers backlog management, plan-to-code traceability, code reviews, and contract validation. Your team gets shared quality standards through house rules.

Enterprise — Everything

specfact init --install all

All six modules. Adds governance enforcement, codebase drift detection, and the full compliance pipeline. Pair with Enterprise Configuration for multi-repo setups.

How Modules Work Together

graph TD
    BL["Backlog<br/><small>Refinement & ceremonies</small>"]
    CR["Code Review<br/><small>Quality scoring & rules</small>"]
    PR["Project<br/><small>Plan-to-code traceability</small>"]
    CB["Codebase<br/><small>Import & drift detection</small>"]
    SP["Spec<br/><small>Contract validation & tests</small>"]
    GV["Govern<br/><small>Quality gate enforcement</small>"]

    BL -- "sync work items" --> PR
    CR -- "review changes" --> PR
    PR -- "analyze code" --> CB
    PR -- "validate contracts" --> SP
    PR -- "enforce gates" --> GV

Backlog feeds work items into Project plans. Project connects plans to code that Codebase analyzes for drift. Spec validates the contracts that Project references. Code Review scores quality on every change. Govern enforces gates before merge and release.


Customization Options by Module

For intermediate and advanced users — each module provides configuration hooks to adapt behavior to your team’s workflow.

Backlog Customizations

What How Use case
Field mapping specfact backlog map-fields or .specfact/backlog-config.yaml Azure DevOps custom process templates with non-standard fields
Adapter selection .specfact/backlog-config.yaml Switch between GitHub Issues and Azure DevOps
Policy framework .specfact/policy.yaml Choose Scrum, Kanban, SAFe, or mixed framework rules
Refinement templates Provider + framework + persona resolution Custom refinement templates per story type (used by CLI and IDE slash prompts)
Ceremony config .specfact/standup.yaml Customize standup format, sprint boundaries, and output

Backlog deep dives → · Custom Field Mapping guide →

Project Customizations

What How Use case
Sync bridges specfact sync bridge config Connect to GitHub, Linear, Jira, OpenSpec, Spec-Kit
Personas specfact project init-personas Map team roles to project bundle access
Import entry points specfact code import --entry-point Specify monorepo roots for brownfield analysis
DevOps stages Stage action configuration Customize the plan → develop → review → release flow

Project deep dives →

Codebase Customizations

What How Use case
Framework extractors Auto-detected (Django, FastAPI, Flask, DRF) Framework-specific feature detection during import
Confidence thresholds CLI flags Tune sensitivity of feature detection
CrossHair depth --crosshair-timeout Control property test exploration time
Sidecar workspace specfact code validate sidecar init Validate external repos without modifying source

Codebase deep dives →

Code Review Customizations

What How Use case
House rules specfact code review rules init Per-IDE quality rules (Cursor, Claude Code, Codex, Windsurf)
Review scope --scope changed vs --scope full Review only changed files or entire codebase
Test inclusion --include-tests / --exclude-tests Control whether test files are reviewed
Score-only mode --score-only CI/CD integration — just the score, no details
Ledger management specfact code review ledger Track quality trends, streaks, and team stats

Code Review deep dives →

Spec Customizations

What How Use case
Validation caching --force-refresh Control when cached validation results expire
Mock server port --port Run multiple mock servers in parallel
Test output directory --output-dir Custom location for generated test suites
SDD constitution specfact spec sdd constitution Customize Software Design Document templates
Generation prompts IDE persona configuration Customize AI-generated contract prompts per IDE

Spec deep dives →

Govern Customizations

What How Use case
Enforcement presets --preset minimal/balanced/strict Choose strictness level for quality gates
SDD manifest path --sdd-manifest Point to custom SDD location
Output format --format yaml/json/markdown Choose report format for CI/CD integration
Non-interactive mode --yes Automated pipelines — skip all prompts
Dry-run patches --dry-run Preview patch effects before applying

Govern deep dives →


Next Steps