# imlazy

> Token-efficient tier-routed agent orchestrator for AI coding assistants. 40+ bundled workflow skills, 19 subagent prompts, and Obsidian vault integration. Works with Claude Code, OpenCode, and Codex CLI.

imlazy classifies every coding task into a tier — Quick, Standard, Heavy, Vault, or Learn — and loads only the workflow weight the task actually warrants. Simple edits cost ~1.5K tokens instead of 14K+.

Key points:
- Quick tier: single-file edits, typo fixes, version bumps, config changes, factual questions — inline execute + verify gate
- Standard tier: 1-3 file features/bugfixes with clear requirements — light planning, TDD, code review
- Heavy tier: multi-component features, ambiguous scope, architectural design — brainstorming, full planning with subagents, TDD, code review
- Vault tier: knowledge persistence across sessions — init vault, save context, search notes
- Learn tier: router misclassification feedback — logged to heuristics.log for adaptive classification
- Zero external plugin dependencies — self-contained, MIT licensed
- Router reads misclassification history to adapt borderline classifications over time

## Docs

- [Homepage](https://hnikoloski.github.io/imlazy/): Project overview, features, quick start, cross-harness support
- [README](https://github.com/hnikoloski/imlazy/blob/main/README.md): Full documentation with tier system, skills table, agents list
- [Install Guide](https://github.com/hnikoloski/imlazy#quick-start): Clone, install, verify in 30 seconds
- [SKILL.md](https://github.com/hnikoloski/imlazy/blob/main/SKILL.md): Entry point orchestrator with tier routing reference

## Skills

- [imlazy-router](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-router/SKILL.md): Task classification and tier dispatch
- [imlazy-bootstrap](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-bootstrap/SKILL.md): Session entry point and trigger detection
- [imlazy-planning](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-planning/SKILL.md): Implementation plans (light and full modes)
- [imlazy-tdd](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-tdd/SKILL.md): Test-driven development with completion gate
- [imlazy-code-review](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-code-review/SKILL.md): Code review request and receive workflow
- [imlazy-brainstorm](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-brainstorm/SKILL.md): Requirements gathering and specification
- [imlazy-debug](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-debug/SKILL.md): Root-cause investigation
- [imlazy-parallel](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-parallel/SKILL.md): Cache-friendly parallel agent dispatch
- [imlazy-worktrees](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-worktrees/SKILL.md): Isolated workspace management
- [imlazy-learn](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-learn/SKILL.md): Router misclassification logging
- [imlazy-vault](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-vault/SKILL.md): Knowledge persistence and retrieval
- [imlazy-wordpress](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-wordpress/SKILL.md): WordPress + ACF development
- [imlazy-security](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-security/SKILL.md): Security review patterns
- [imlazy-api-design](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-api-design/SKILL.md): REST API design
- [imlazy-frontend](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-frontend/SKILL.md): Frontend development
- [imlazy-backend](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-backend/SKILL.md): Backend architecture
- [imlazy-standards](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-standards/SKILL.md): Coding conventions
- [imlazy-verify](https://github.com/hnikoloski/imlazy/blob/main/skills/imlazy-verify/SKILL.md): Pre-completion verification

## Subagents

- [imlazy-planner](https://github.com/hnikoloski/imlazy/blob/main/agents/imlazy-planner.md): Implementation planning
- [imlazy-architect](https://github.com/hnikoloski/imlazy/blob/main/agents/imlazy-architect.md): System architecture
- [imlazy-code-reviewer](https://github.com/hnikoloski/imlazy/blob/main/agents/imlazy-code-reviewer.md): Code review
- [imlazy-security-reviewer](https://github.com/hnikoloski/imlazy/blob/main/agents/imlazy-security-reviewer.md): Security review
- [imlazy-tdd](https://github.com/hnikoloski/imlazy/blob/main/agents/imlazy-tdd.md): TDD coaching
- [imlazy-build-fixer](https://github.com/hnikoloski/imlazy/blob/main/agents/imlazy-build-fixer.md): Build error resolution
- Language reviewers: Python, Go, Rust, Java, C++, Kotlin, PHP

## Optional

- [GitHub Repository](https://github.com/hnikoloski/imlazy): Source code, issues, pull requests
- [LICENSE](https://github.com/hnikoloski/imlazy/blob/main/LICENSE): MIT License
- [ATTRIBUTIONS](https://github.com/hnikoloski/imlazy/blob/main/ATTRIBUTIONS.md): Third-party acknowledgments