session-orchestrator
Structured coding sessions with wave execution and quality gates
All Skills in This Repository (13)
Pure Emerald Level Indicatorssession-start
Initialize and analyze git repository state and project readiness for engineering sessions.
bootstrap
Scaffold minimal repository structures with CLAUDE.md and AGENTS.md files.
wave-executor
Coordinate multi-agent engineering sessions into structured waves with quality checks.
quality-gates
Run typecheck, test, and lint quality gates across agent-driven sessions.
plan
Generate structured PRDs and prioritized issue plans for software projects.
ecosystem-health
Monitors service health, CI pipelines, and critical issues across repositories.
session-end
Finalize engineering sessions by verifying plans, running quality gates, and committing changes.
discovery
Detect and verify codebase issues with automated probes and confidence scoring.
vault-sync
Validate Markdown frontmatter and wiki-link integrity in Obsidian vaults.
evolve
Extract session patterns from JSONL metrics into structured learnings.
daily
Create today's daily note with vault-compliant YAML frontmatter.
gitlab-ops
Standardize GitLab and GitHub CLI usage with dynamic project resolution.
Frequently Asked Questions
FAQPage SchemaHow to install session-orchestrator?βΌ
Run `npx skills add Kanevry/session-orchestrator --all -g -y` in your terminal to install all skills globally. Node.js 24 or later is required.
What does session-orchestrator do?βΌ
It structures coding work into a repeatable loop: research and scope alignment, then execution in five typed waves of parallel subagents with quality gates between them, then a verified close-out with clean commits.
Which coding assistants does it work with?βΌ
The same skills and commands run on Claude Code, Codex CLI, Cursor IDE, and Pi, with platform-adapted hooks and enforcement on each.
How do I run a typical session?βΌ
Run /bootstrap once in your repo, then use three commands: /session to research and agree on scope, /go to execute the wave plan, and /close to verify, commit, and file carryover issues.
Does it prevent agents from making unsafe changes?βΌ
Yes. Hooks block destructive shell commands and out-of-scope file edits, circuit breakers stop runaway agents, and STATE.md persists progress so a crashed session can resume safely.
Related Repositories in Software Engineering
View All in Software Engineeringβopenclaw
Run a personal AI assistant across your devices and chat apps
superpowers
Gives coding agents a disciplined workflow from idea to merged code
react
AI agent skills for building, testing, and porting React core