Elevate your AI coding sessions.
A centralized suite of security hooks, declarative skills, and agent profiles. Cross-compatible with Claude Code and Antigravity (Gemini).
End-to-End Guidance
A specialized AI companion for every stage of your SDLC. From scaffolding architecture and planning commits, to reviewing PRs and drafting semantic releases.
Declarative Skills
A portable engine of reusable scripts written in Markdown. Tell the AI exactly how to review PRs, draft releases, or scaffold architecture.
Auto-Documentation
Never lose context. Automated session summaries record goals, decisions, file changes, and pending tasks behind the scenes.
Suite Capabilities
A complete ecosystem of background hooks, autonomous agents, and declarative skills.
Autonomous Agents
ami-commit-assistant
Interactively structures, formats, and executes git commits securely.
ami-push-assistant
Validates tests and code quality before pushing to remote repositories.
ami-pr-publisher
Automates the creation, description, and formatting of Pull Requests.
ami-next-step-assistant
Analyzes current repository state and proposes logical next actions.
Declarative Skills
ami-commit-planner
Analyzes uncommitted work and proposes a structured set of conventional commits.
ami-context-researcher
Actively researches external documentation and saves findings to prevent context loss.
ami-data-validator
Validates structural consistency between code changes and database schemas.
ami-doc-architect
Scans the repo to understand its structure and interactively builds project docs.
ami-docs-updater
Identifies and updates existing documentation to reflect recent code changes.
ami-learnings-extractor
Extracts architectural decisions, lessons, and patterns from recent code changes.
ami-pr-comment-analyzer
Analyzes PR review comments, extracts pending tasks, and helps you reply.
ami-pr-conflict-detector
Alerts if parallel Pull Requests might conflict with your current working branch.
ami-pr-peer-reviewer
Assists in reviewing others' Pull Requests, focusing on code quality and logic.
ami-pr-self-reviewer
Acts as a critical self-reviewer for your own Pull Requests and proactively suggests concrete code fixes.
ami-project-architect
Interactively scaffolds the initial architecture, stack, and folders of a new project.
ami-quality-auditor
Performs a deep code quality, security, and structure audit prior to a PR.
ami-quick-reviewer
Static analysis of modified files before a commit to ensure English-only rules.
ami-release-drafter
Analyzes git commits to automatically draft bilingual semantic release notes.
ami-session-summarizer
Summarizes the current AI session using internal memory and saves it as a structured Markdown report.
ami-tech-debt-scanner
Scans the repository for technical debt, outdated dependencies, and dead code.
ami-test-creator
Automatically generates unit tests for modified code if none exist.
ami-test-runner
Finds and executes the local test suite to ensure no regressions before a PR.
Background Hooks
Session Context Restore
Injects pending tasks from previous sessions automatically on startup.
Git Pre-Flight Guards
Blocks raw CLI git commands to enforce AI assistant review workflows.
Real-time Debug Linter
Scans modified files instantly for leftover console.log or TODO statements.
Quick Start
Global Installation
Uses the CLI wizard to merge the background hooks directly into your system's global settings.json.