Component Catalog

Full listing of all commands, agents, skills, hooks, and MCP servers included in Lavra.

See the Command Reference for descriptions and direct links, or the Command Map for a visual overview.

Commands (23 core + 5 optional)

Workflow (12)

CommandDescriptionWhen to Use
/lavra-designOrchestrate the full design pipeline, brainstorm, plan, research, revise, review, lockRuns the entire design pipeline automatically
/lavra-brainstormExplore requirements and approaches through collaborative dialogue before planningWhen requirements are unclear or you need to explore approaches. Included in /lavra-design
/lavra-planTransform feature descriptions into well-structured beads with parallel research and multi-phase planningStart every feature, creates a structured plan with research. Included in /lavra-design
/lavra-reviewPerform exhaustive code reviews using multi-agent analysis and ultra-thinkingBefore closing beads, comprehensive quality check. Included in /lavra-design
/lavra-quickFast-track small tasks, abbreviated plan then straight to executionQuick fixes and small features that don’t need the full pipeline
/lavra-workExecute work on one or many beads, auto-routes between single-bead and multi-bead paths based on inputStandard workflow, any number of beads
/lavra-work-ralphAutonomous retry mode, iterates until completion criteria are met or retry budget is exhaustedHands-off execution with self-correction
/lavra-work-teamsWork on multiple beads with persistent worker teammates that self-organize through a ready queueSpeed up delivery with team parallelism
/lavra-qaBrowser-based QA verification of the running app, systematic testing from the user’s perspectiveAfter implementation, verify the feature works end to end
/lavra-shipFully automated ship sequence from code-ready to PR-open with beads closed and knowledge capturedWhen work is done, opens PR, closes beads, pushes
/lavra-checkpointSave session progress by filing beads, capturing knowledge, and syncing stateMid-session, checkpoint your work before context runs out
/lavra-retroWeekly retrospective with shipping analytics, team performance, and knowledge synthesisWeekly, review what shipped and surface deferred decisions

Planning & Triage (5)

CommandDescriptionWhen to Use
/lavra-researchGather evidence and best practices for a plan using domain-matched research agentsFor complex features. Gathers best practices and prior art. Included in /lavra-design
/lavra-ceo-reviewCEO/founder-mode plan review, challenge premises, validate business fit, run 10-section structured reviewBefore committing to a feature, validate scope and business fit. Included in /lavra-design
/lavra-eng-reviewEngineering review, parallel agents check architecture, simplicity, security, and performanceBefore implementation, catch design issues early. Included in /lavra-design
/lavra-triageTriage and categorize beads for prioritizationAfter planning or review, organize the work queue
/lavra-importImport a markdown plan into beads as an epic with child tasksWhen you have an external plan to convert into beads

Knowledge (2)

CommandDescriptionWhen to Use
/lavra-learnCurate raw knowledge comments into structured, well-tagged entries for future auto-recallAfter closing beads, clean up terse knowledge captured during work
/lavra-recallSearch knowledge base mid-session and inject relevant contextWhen you need past learnings mid-session without restarting

Utility (4)

CommandDescriptionWhen to Use
/changelogCreate engaging changelogs for recent merges to main branchBefore a release, summarize what shipped
/heal-skillFix incorrect SKILL.md files when a skill has wrong instructions or outdated API referencesWhen a skill is misbehaving or referencing stale APIs
/test-browserRun browser tests on pages affected by current PR or branchAfter UI changes, verify nothing broke in the browser
/report-bugReport a bug in the Lavra pluginWhen you hit a bug in Lavra itself

Optional (5)

Domain-specific commands in commands/optional/. Not installed by default, copy manually to .claude/commands/ to enable.

CommandDescriptionWhen to Use
/feature-videoRecord a video walkthrough of a feature and add it to the PR descriptionWhen a PR benefits from a visual demo
/agent-native-auditRun comprehensive agent-native architecture review with scored principlesWhen designing systems where agents are first-class users
/xcode-testBuild and test iOS apps on simulator using XcodeBuildMCPiOS projects, run tests without leaving the agent
/reproduce-bugReproduce and investigate a bug using logs, console inspection, and browser screenshotsBefore filing a bug, confirm the issue is reproducible
/generate-commandCreate a new custom slash command following conventions and best practicesWhen you want to extend Lavra with a project-specific command

Agents (30): Cost-Optimized by Model Tier

Haiku class model (5): Structured tasks, fast and cheap

learnings-researcher, repo-research-analyst, framework-docs-researcher, ankane-readme-writer, lint

Sonnet class model (18): Moderate judgment, balanced cost

best-practices-researcher, bug-reproduction-validator, code-simplicity-reviewer, deployment-verification-agent, design-implementation-reviewer, design-iterator, dhh-rails-reviewer, every-style-editor, figma-design-sync, git-history-analyzer, goal-verifier, kieran-python-reviewer, kieran-rails-reviewer, kieran-typescript-reviewer, migration-drift-detector, pattern-recognition-specialist, pr-comment-resolver, security-sentinel

Opus class model (7): Deep reasoning, premium quality

agent-native-reviewer, architecture-strategist, data-integrity-guardian, data-migration-expert, julik-frontend-races-reviewer, performance-oracle, spec-flow-analyzer

Skills (15)

Core (8): Installed by default

SkillDescription
agent-browserBrowser automation for testing and screenshots
agent-native-architectureDesign agent-native system architectures
brainstormingStructured brainstorming with bead output
create-agent-skillsCreate new agents and skills
file-todosFind and manage TODO comments in code
git-worktreeManage git worktrees for parallel bead work
lavra-knowledgeDocument solved problems as knowledge entries
project-setupProject stack detection, review agent config, and optional codebase analysis

Optional (7): Copy from skills/optional/ to use

SkillDescription
andrew-kane-gem-writerWrite Ruby gems following Andrew Kane’s style
dhh-rails-styleRails development following DHH’s conventions
dspy-rubyDSPy integration for Ruby applications
every-style-editorEvery’s house style guide for content editing
frontend-designFrontend design patterns and best practices
gemini-imagegenGenerate images using Google’s Gemini
rcloneCloud storage file management with rclone

MCP Servers

  • Context7: Framework documentation lookup

Hooks (4)

HookTriggerPurpose
auto-recall.shSessionStartInject relevant knowledge at session start
memory-capture.shPostToolUse (Bash)Extract knowledge from bd comment calls
subagent-wrapup.shSubagentStopEnsure subagents log learnings before completing
teammate-idle-check.shTeammateIdlePrevent --teams workers from idling while ready beads remain