Command Reference¶
Blueprint provides 43 commands organized into 8 categories, covering the full lifecycle of architectural governance: setup, decision management, analysis, evaluation, documentation, continuous enforcement, strategic planning, and system administration.
Essential Commands¶
If you read nothing else, learn these three:
| Command | What it does |
|---|---|
/blueprint:init |
Scan your codebase and bootstrap the ADR system |
/blueprint:new "topic" |
Record an architectural decision |
/blueprint:review N |
Challenge a decision before accepting it |
Categories¶
| Category | Commands | Page |
|---|---|---|
| Setup | init, quickstart, onboard, hooks | Setup |
| Lifecycle | advise, new, list, review, challenge, transition, search, grill-me, help, rearchitect | Lifecycle |
| Analysis | impact, audit, reflect, evidence, map, tradeoff, risk, trace, retro | Analysis |
| Evaluation Team | evaluate (5 dimensions: consistency, bugs, maintainability, testing, conways) | Evaluation |
| Documentation | architect, diagram, eli5, digest, timeline, export, views | Documentation |
| Continuous Governance | fitness, drift, debt, guard, nudge | Continuous Governance |
| Strategic | scope, radar, govern, federate | Strategic |
| System | status, health | System |
Command Anatomy¶
Every Blueprint command follows a consistent pattern:
- Positional arguments are typically ADR numbers (
3), topics ("use PostgreSQL"), or subcommands (accept). - Flags modify behavior (
--research,--format github-actions,--flat). - All commands auto-detect the ADR directory using the config resolution protocol.
- Read-only commands (list, search, status, trace) never modify files.