Orchestkit
Free MIT Claude Code plugin with 110 skills, 36 agents, and 162 lifecycle hooks that block dangerous commands.
The hooks are the reason to install. 162 lifecycle gates that stop destructive commands before they run do more for agent safety than any prompt instruction, and the October beta added a review-only standards file so /ork:review-pr grades against your conventions instead of generic advice. The catch is that all of it only registers inside Claude Code 2.1.277+. Pick it if Claude Code is already your daily driver.
Verified 10h ago · liveness 69/100 · cite: rightaichoice.com/tools/orchestkit
- Claude Code 2.1.277+ users who want security and quality gates enforced on every tool call, not suggested in a prompt
- Teams standardizing PR review and parallel security audits across an existing codebase
- Engineers who want a searchable catalog of 110 skills instead of hand-writing .claude/ config from scratch
- Tech leads tired of re-explaining stack conventions to the agent at the start of every session
- Anyone not using Claude Code — the hook layer only registers there, so you lose the guardrails
- Developers shopping for IDE autocomplete like Cursor or GitHub Copilot — this is an agent-layer tool
- Teams that need a frozen release line — the current build is a frequent 10.0.0 beta with v9.8.1 as stable
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip OrchestKit if you are not running Claude Code 2.1.277+ — the 169 guardrail hooks only register there, so on Cursor, Codex, Devin, OpenCode or Antigravity you get a reduced skill set with none of the safety gates that justify the install.
OrchestKit itself costs $0, but you still need Claude Code and an Anthropic account — that billing is between you and Anthropic and is entirely independent of the plugin
OrchestKit is $0 under the MIT license — no paid tiers, no usage limits, no account — so pricing is never the deciding factor. What you actually compare is your existing Anthropic Claude Code billing, which OrchestKit does not touch. Against paid agent-layer tools with per-seat subscriptions, OrchestKit costs strictly less; the tradeoff is the beta release line and the Claude Code-only hook layer.
In short
Orchestkit — Free MIT Claude Code plugin with 110 skills, 36 agents, and 162 lifecycle hooks that block dangerous commands. Best for Claude Code 2.1.277+ users who want security and quality gates enforced on every tool call, not suggested in a prompt, Teams standardizing PR review and parallel security audits across an existing codebase, Engineers who want a searchable catalog of 110 skills instead of hand-writing .claude/ config from scratch. Free to use.
What's new in Orchestkit
Checked 6 days agoAcross the latest 5 updates: 3 feature updates and 2 changelog entries.
review-pr: standards pass with a review-only standards file (#4671)
Adds a review-only standards file so the /ork:review-pr pass grades against your project standards. The same release bounds the CI cache on the docs Playwright install and restores the exec bit on git-isolate.sh.
prompt-focus: where your attention went, from your own prompts (#4657)
Ships a new prompt-focus primitive that traces where your attention went, derived from your own prompts.
hooks: display-lint hint is shell-aware, model-only, once per session (#4655)
A lifecycle hook changed: the display-lint hint is now shell-aware, model-only and fires once per session. Skills also load inline and ork:auto routes end-to-end asks.
glyph: output mode dials, done sign-off, and a try-it page (#4647)
Adds glyph output mode dials and a done sign-off, plus a try-it page. Also bumps the vitest dev dependency group.
deps: sharp 0.35.5 in docs/site for advisory 1241331 (#4653)
Upgrades sharp to 0.35.5 in docs/site to resolve advisory 1241331.
Viability Score
How well maintained and how widely used is Orchestkit? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- 110 reusable skills searchable by category across development, frontend, AI, backend, testing, security, DevOps,
- 36 specialist agents that fan out in parallel across isolated git worktrees
- 162 TypeScript lifecycle hooks that block dangerous commands before execution
- Non-blocking hook design for context injection and security enforcement without stalling the loop
- 3-tier memory: knowledge graph, local persistence and MEMORY.md sync across sessions
- memory-lens mod that recalls related memories on each prompt
- prompt-focus primitive that traces where your attention went, derived from your own prompts (#4657)
- /ork:review-pr grades against a review-only standards file and posts one synthesized comment (#4671)
- Eight recipes including /ork:implement, /ork:design-ship, /ork:review-pr, /ork:fix-issue and /ork:audit-full
- /ork:audit-full scans auth, secrets, OWASP and dependencies in parallel
- Demo video pipeline via VHS + Remotion with auto-generated voiceover
- Multi-agent task management with TaskCreate / TaskList and dependency chains
- Agent-readable docs: append .md to any URL, or send Accept: text/markdown for raw Markdown
- MCP server over Streamable HTTP at /api/mcp with documentation search and Markdown page fetch
- NLWeb /ask endpoint for natural-language doc queries with SSE streaming
About Orchestkit
OrchestKit is a free, MIT-licensed plugin library for Claude Code, Anthropic's agentic command-line coding tool, and it wires security gates and review standards directly into the agent loop. One marketplace install drops 110 skills, 36 specialist agents, and 162 lifecycle hooks into your project, so Claude Code starts from your conventions instead of a blank prompt. Current build is v10.0.0-beta.182, the repo sits at roughly 296 GitHub stars, and it requires Claude Code 2.1.277+. The three layers have distinct jobs. Hooks fire automatically on every action (security gates, auto-lint hints, memory injection) and are never called by hand. Agents activate from keywords and fan out in parallel across isolated git worktrees. Skills load on demand into whichever agent needs them, indexed across development, frontend, AI, backend, testing, security, DevOps, product, data and research categories. Eight cookbook recipes turn that into runnable workflows: /ork:implement (idea to merged PR), /ork:design-ship (URL in, reviewable PR out), /ork:review-pr (parallel reviewers, one synthesized comment), /ork:fix-issue, /ork:audit-full (parallel scan across auth, secrets, OWASP, dependencies), /ork:memory, /ork:demo-producer and multi-agent task management. Recent 10.0.0 betas added a review-only standards file for review-pr (#4671) and a prompt-focus primitive that traces where your attention went from your own prompts (#4657). The docs are written for agents too: append .md to any URL, start at /llms.txt, read the OpenAPI spec at /openapi.json, query the Streamable HTTP MCP server at /api/mcp, or hit the NLWeb /ask endpoint. The same surface ships as npm, PyPI and Go SDK packages. On Cursor, Codex, Devin and OpenCode you get a reduced skill set without the hook layer, so OrchestKit sits in the agent layer rather than competing with your editor.
Behind the Verdict
Most 'AI coding' tooling competes for the editor. OrchestKit takes the other seat, making the agent inside Claude Code behave more like a senior teammate. In practice the difference shows up in the boring moments: the tool call that would have dropped a table gets blocked, the lint hint arrives once rather than five times per session. Pick this when Claude Code is already where your team works and you are tired of re-pasting stack conventions at the top of every session. The install is two commands and /ork:setup; there is no account, no seat licensing and no telemetry, which makes it an easy sell to a security-minded lead. Pass if your team lives in Cursor, Codex or Devin. You can still install skills there, but the hook layer only registers on Claude Code, so the guardrails — the actual reason to bother — do not come with you. Also pass if you need a frozen release line: the current build is a 10.0.0 beta with frequent point releases, and the stable channel is pinned to v9.8.1. The closest thing to a substitute is hand-rolling your own .claude/ config with a few CLAUDE.md rules. That works until you want 36 agents and 169-style hook coverage maintained by someone else; the tradeoff is you inherit beta churn. Where it bites: a standards gate cannot retroactively catch a session already open in CI, and recipes assume your repo roughly matches the git-worktree flow they were written for. We'd reach for OrchestKit as the first thing we install in a new Claude Code project, then strip recipes and skills we don't use. What you pay for separately is the model itself — OrchestKit is free, Anthropic billing is not.
Researching Orchestkit? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas Orchestkit actually fits — and what changes day-one when you adopt it.
Installs via 'claude plugin marketplace add yonatangross/orchestkit' then 'claude plugin install ork@orchestkit', runs /ork:setup, and describes a feature instead of pasting stack conventions. Hooks fire on every tool call and /ork:commit writes conventional commits.
Outcome: First value in about 2 minutes; you stop re-explaining FastAPI/SQLAlchemy patterns and OWASP checks in every prompt.
Points /ork:review-pr at an open PR. Three specialized reviewer agents spawn in parallel across isolated worktrees and grade against a review-only standards file shipped in 10.0.0-beta.166.
Outcome: You get one synthesized review comment instead of five overlapping ones, and the review reflects your conventions rather than generic advice.
Runs /ork:audit-full, which scans auth, secrets, OWASP categories and dependencies in parallel rather than as a serial checklist.
Outcome: A pre-release security pass that runs alongside normal agent work, with 169 lifecycle hooks still gating destructive commands as you go.
Use Cases
- Implement a feature from idea to merged PR with parallel AI agents using /ork:implement
- Review a pull request with parallel specialized reviewers grading against your own standards file using /ork:review-pr
- Fix a GitHub issue with root-cause analysis, regression detection and a linked PR using /ork:fix-issue
- Run a comprehensive security audit across auth, secrets, OWASP and dependencies in parallel using /ork:audit-full
- Design new features and automatically generate a reviewable PR from a URL using /ork:design-ship
- Set up cross-session memory with a 3-tier knowledge graph using /ork:memory
- Produce a demo video with a VHS + Remotion pipeline and auto-generated voiceover using /ork:demo-producer
- Manage multi-agent work with TaskCreate, TaskList and dependency chains
Models Under the Hood
as of 2026-10-18
Limitations
- OrchestKit is a free, open-source (MIT) plugin that runs locally inside Claude Code, Anthropic's agentic command-line coding tool; it is not a hosted service and not a general-purpose editor assistant.
- It requires Claude Code 2.1.277+ and a separate Anthropic account, whose billing is between the user and Anthropic and independent of OrchestKit.
- The current release line is v10.0.0-beta.182 (beta) out of 690 releases, indicating rapid beta-stage development, so teams that need a frozen release line should be aware of that pace.
- OrchestKit is free with no paid tiers, seat licensing, or usage limits.
as of 2026-10-12
Verification history
We have re-verified Orchestkit 10 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Showing the 6 most recent of 10 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published Orchestkit tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
OrchestKit (MIT)
$0
Ideal for
Any developer or team already running Claude Code 2.1.277+ who wants the 169 guardrail hooks and the 110-skill catalog without a paid tier or seat license
What this tier adds
Starting tier and only tier — $0 under MIT, no usage limits, no account, no telemetry
Where the pricing makes sense
The company stage and team size where Orchestkit's pricing actually pencils out — and where peers do it cheaper.
OrchestKit is $0 under the MIT license — no paid tiers, no usage limits, no account — so pricing is never the deciding factor. What you actually compare is your existing Anthropic Claude Code billing, which OrchestKit does not touch. Against paid agent-layer tools with per-seat subscriptions, OrchestKit costs strictly less; the tradeoff is the beta release line and the Claude Code-only hook layer.
Setup time & first value
How long it actually takes to get something useful out of Orchestkit — broken out by persona, not the marketing-page minute.
About 2 minutes if you already run Claude Code 2.1.277+: add the marketplace, install ork@orchestkit, then /ork:setup for the guided first session. Teams adding it to an existing codebase should budget an extra pass to encode stack conventions into skills and the review standards file. If your team is partly on Cursor, Codex or Devin, add time for the split setup, since hooks only register on
Switching to or from Orchestkit
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From hand-written .claude/ config: replace your own skill and agent files with the marketplace catalog's 110 skills and 36 agents, then keep repo-specific rules as overrides
- →From a bespoke prompt library: move recurring stack and review instructions into OrchestKit skills so they load on demand instead of being pasted each session
- →From other Claude Code plugins: verify hook conflicts before installing, since 169 lifecycle hooks gate every tool call
- →From ad-hoc PR review checklists: adopt /ork:review-pr with a review-only standards file so reviewers grade against your conventions
- ↗To plain Claude Code: uninstall the plugin and re-encode whichever skills and hook rules you still need by hand in .claude/
- ↗To Cursor or another agent: install the Cursor plugin or Agent Skills route for the skill set, accepting that the 169 guardrail hooks do not register
- ↗To an IDE autocomplete tool: nothing carries over, because OrchestKit is an agent-layer tool and does not do inline completion
Integrations
Resources & Guides
- Documentationorchestkit.yonyon.ai
Docs · Orchestkit
Full product docs from orchestkit.yonyon.ai
- Quickstartorchestkit.yonyon.ai
Getting Started · Orchestkit
Get up and running fast from orchestkit.yonyon.ai
- API Referenceorchestkit.yonyon.ai
Sdk · Orchestkit
Methods, params, types from orchestkit.yonyon.ai
- Documentationorchestkit.yonyon.ai
Mcp · Orchestkit
Full product docs from orchestkit.yonyon.ai
- Resourceorchestkit.yonyon.ai
Changelog · Orchestkit
Helpful link from orchestkit.yonyon.ai
- Resourceorchestkit.yonyon.ai
Openapi · Orchestkit
Helpful link from orchestkit.yonyon.ai
- Resourceorchestkit.yonyon.ai
Llms · Orchestkit
Helpful link from orchestkit.yonyon.ai
- Resourceorchestkit.yonyon.ai
Pricing · Orchestkit
Helpful link from orchestkit.yonyon.ai
Tutorials & Learning
YouTube returned 6 videos for “Orchestkit”, and we withheld 6: 6 could not be judged, because “Orchestkit” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about Orchestkit.
Official links
Tools that pair well with Orchestkit
Common stack mates teams adopt alongside Orchestkit, with the specific reason each pairing earns its keep.
Ellipsis
Run Claude Code and Codex coding agents in managed cloud sandboxes defined by YAML in your repo.
Continue
Open-source AI coding agent for VS Code and JetBrains — free to fork since Cursor acquired Continue in January 2026.
Warp
Warp Factories runs fleets of cloud coding agents across your SDLC, defined in code as factory.yaml.
Featured Head-to-Head Comparisons
Orchestkit vs Spider Cloud
Spider Cloud and Orchestkit serve entirely different needs. Spider Cloud is a web crawling & scraping API for AI data ingestion (RAG, LLM context). Orchestkit is a free plugin that supercharges Claude Code for AI-assisted development. Pick Spider Cloud if you need to pull real-time web data at scale; pick Orchestkit if you already use Claude Code and want automated security, memory, and multi-agent workflows. They are not direct competitors.
Orchestkit vs Voyage Ai
Voyage AI and OrchestKit serve completely different needs: Voyage is a paid, enterprise-grade embedding/reranker service for high-accuracy RAG on domain-specific data, while OrchestKit is a free, open-source plugin that supercharges Claude Code with 113 skills, 37 agents, and 212 hooks for automated security and quality. Pick Voyage if you need a compliant, high-performance retrieval backend for complex documents; pick OrchestKit if you already use Claude Code and want agentic guardrails and reusable workflows.
Orchestkit vs Temporal Ai
Temporal AI and Orchestkit serve fundamentally different needs. Temporal is a full-platform durable execution engine for building reliable, long-running AI workflows and microservices — it's overkill for simple tasks but essential for mission-critical systems. Orchestkit is a focused, free plugin that supercharges Claude Code with reusable skills, specialist agents, and automated guardrails, ideal for teams already using Claude Code who want to enforce quality and speed up reviews without leaving the CLI. Choose Temporal if you need crash-proof orchestration across services; choose Orchestkit if you live in Claude Code and want smarter, safer coding.
Marvin vs Orchestkit
Marvin is the right choice if you're a Python developer who needs to integrate LLMs into your application code with type safety and minimal overhead. Orchestkit is the clear winner if you already use Claude Code and want to supercharge it with reusable skills, parallel agents, and automated guardrails without context loss. Your choice depends entirely on whether you're building Python-first LLM apps or enhancing an existing Claude Code workflow.
Alternatives to Orchestkit
View all
Frequently Asked Questions
Best-of guides
Used Orchestkit? Help shape our editorial sentiment research.