Registry
The ACP Registry is one JSON endpoint listing every ACP-compatible coding agent you can install and authenticate.
For ACP client developers, the registry is the shortest path from an empty agent picker to a populated one, and the authentication requirement does real filtering work for you. The trade-off is that it ships metadata, not judgment: nothing in the catalog tells you whether an agent is any good, so you still run your own evaluations. Treat it as a distribution feed, not a review site.
Verified 5d ago · liveness 73/100 · cite: rightaichoice.com/tools/registry
- Client developers who need a ready-made ACP agent directory instead of hand-tracking vendor repos
- IDE and editor teams wiring an install-and-launch picker into an ACP-speaking tool
- Agent vendors who want a reviewed, versioned distribution channel with authentication as the bar
- Security-conscious teams that want authentication as a hard listing requirement
- Non-technical users expecting a graphical app store with ratings, reviews, or screenshots
- Agents with no authentication flow, which the registry will not list
- Projects that do not speak the Agent Client Protocol
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 the ACP Registry if your stack does not speak the Agent Client Protocol, or if you need an app store with ratings, reviews or vendor support commitments before you install anything.
The registry itself is free, but each listed agent runs on its own vendor's paid plan or API billing, so your real spend lands with Amp, Claude, Cursor or whichever agent you pick.
The registry itself is a free, read-only JSON catalog fetched from the CDN — no account needed to browse or pull it. The cost question sits downstream: each listed agent carries its own vendor pricing, from open-source CLI agents such as goose, Kilo and OpenCode that you self-fund, to commercial agents like Cursor, Devin and GitHub Copilot with their own seat and usage billing.
In short
Registry — The ACP Registry is one JSON endpoint listing every ACP-compatible coding agent you can install and authenticate. Best for Client developers who need a ready-made ACP agent directory instead of hand-tracking vendor repos, IDE and editor teams wiring an install-and-launch picker into an ACP-speaking tool, Agent vendors who want a reviewed, versioned distribution channel with authentication as the bar. Free to use.
What's new in Registry
Checked 4 days agoAcross the latest 5 updates: 3 feature updates, 1 launch and 1 changelog entry.
Tool Call Names stabilized in ACP
The Tool Call Name RFD completed, making optional programmatic tool names stable in ACP v1 and the v2 baseline so tool identity is separate from the display title.
Elicitation stabilized in ACP
Structured form and URL-mode user interactions are now stabilized, with clients advertising which elicitation modes they support.
ACP v2 published in draft
The ACP v2 protocol documentation and schema were released in Draft form as the next version of the Agent Client Protocol.
Request cancellation stabilized in ACP
The Request Cancellation RFD completed, stabilizing the $/cancel_request notification so either side can cancel an outstanding JSON-RPC request by ID.
Rust and TypeScript ACP SDKs reach 1.0
The Rust and TypeScript SDKs hit 1.0.0, giving stable foundations for building ACP clients, agents and protocol adapters.
What people actually say about Registry — is it worth it?
We scanned public community sources for Registry on Jul 3, 2026 and could not establish that the discussion we found is about this tool rather than something else sharing its name. Our own analysis of that scan says the posts were off-subject. Rather than publish a sentiment score built on the wrong subject, we publish nothing here and re-run the scan.
Viability Score
How well maintained and how widely used is Registry? 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
- Curated catalog of ACP-compatible coding, DevOps, and general-purpose agents
- Every listed agent must support authentication to be admitted
- Single CDN endpoint serves the whole catalog as registry JSON
- Per-agent versioned metadata with exact version numbers
- Distribution info enabling automated installation by ACP clients
- Optional icon.svg per listing, 16x16 recommended
- Fork-based submission via GitHub pull request
- agent.json schema for standardized agent submissions
- Maintainer review for authentication compliance and schema fit
- Read-only catalog requiring no account or API key
- Covers agents from Anthropic, OpenAI, Google, xAI, Mistral, Alibaba, JetBrains, Snowflake and Cognition
- Includes DevOps agents such as Stakpak and general-purpose agents such as LangChain-powered DeepAgents
- Individual agents expose streaming, tool calls, image input, mid-session model switching, and session load/fork/resume
- ACP v1 stable with v2 available in draft
- Local-first and marketplace listings coexist, including Agoragentic and siGit Code
About Registry
If you maintain an agent picker inside an ACP client, the ACP Registry saves you from hand-tracking a few dozen vendor repos. One request to https://cdn.agentclientprotocol.com/registry/v1/latest/registry.json returns the full catalog as structured registry JSON: agent ID, version number, an optional 16x16 icon, and the distribution metadata a client needs to install and launch the agent. Builders of clients, IDE plugins, and internal dev tooling are the people this is actually for. The catalog skews hard toward coding agents, and it has grown well past the vendor list most teams would guess. Coding entries include Amp, Auggie CLI, Claude Agent, Cline, Codebuddy Code, Codex, Copilot, Cursor, Devin, Factory Droid, Gemini CLI, Google Antigravity, goose, Grok Build, JetBrains Junie, Kilo, Kimi CLI, MiniMax Code, Mistral Vibe, OpenCode, Poolside, Qoder CLI and Qwen Code. Snowflake's Cortex Code and Z.ai's GLM Agent sit alongside general-purpose and DevOps entries such as DeepAgents and Stakpak. The admissions rule is the part worth reading twice: every listing must support authentication. That is the bar before an agent is handed credentials to your repository, and it removes agents that would take repo access without a login step. Z.ai's GLM Agent listing, for example, documents streaming, tool calls, mid-session model switching, image input via the Coding Plan Vision MCP, and session load/fork/resume with on-disk persistence. Publishing your own agent is a normal GitHub flow: fork the repository, create a folder named for your lowercase hyphenated agent ID, add an agent.json against the schema, optionally an icon.svg, and open a pull request. Maintainers review for authentication compliance and schema fit, not for taste. Outside ACP this is irrelevant, and even inside ACP it is an index rather than a product: no quality benchmarks, no ratings, no support contract.
Behind the Verdict
The honest framing: the ACP Registry is infrastructure, not a destination. You will never "browse" it the way you browse an app store. You curl it, parse it, and render it inside whatever client you are building, and that is the entire intended experience. Pick it up when you are writing an ACP client or an IDE integration and you need install metadata for agents you did not write. Version pinning matters here. Each listing carries an exact version, so a client can resolve a specific release rather than floating on whatever main happens to be. The protocol itself has been settling down too: the Rust and TypeScript SDKs hit 1.0.0 in June 2026, and through mid-2026 the RFDs for request cancellation, elicitation, and tool call names all stabilized, which means the surface you build against is less likely to shift under you. ACP v2 is published in draft, so plan for a second version of the schema at some point. Pass on it if your toolchain has nothing to do with ACP, if you want a graphical storefront with star ratings and user reviews, or if your procurement process requires a vendor SLA. There is no commercial escalation path here, and nobody is on call for your integration. The closest alternative is simply maintaining your own list of agent repos and launch commands. That works, and plenty of teams do it, but it goes stale the moment a vendor renames a binary or ships a new install route. The registry centralizes that maintenance onto the agent authors themselves. One practical caveat: authentication being mandatory is a floor, not a security review. It tells you an agent has a login step. It does not tell you what that agent does with credentials afterward, what it sends to which model provider, or whether it phones home. Do that homework separately, per agent,
Researching Registry? 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 Registry actually fits — and what changes day-one when you adopt it.
You curl the registry endpoint from your build or first-run path, parse the JSON, and render the agent list with versions and icons already attached instead of writing per-vendor launch logic.
Outcome: You ship an agent picker in an afternoon rather than maintaining a hand-curated list of vendor install instructions.
You fork the registry repository, add a folder named for your lowercase hyphenated agent ID, write agent.json against the schema, add an icon.svg, and open a pull request.
Outcome: Your agent appears in the catalog every ACP-speaking client already reads, without negotiating a separate distribution deal with each client.
You pull the registry JSON, compare pinned versions for Claude Agent, Codex, Cursor and Gemini CLI, and pick one agent per workflow, with authentication as a built-in requirement rather than an audit you run yourself.
Outcome: A shortlist built from current versioned metadata instead of stale README pages, with the credential-handling question already answered by the listing rule.
Use Cases
- Discover and install ACP-compatible coding agents for your development workflow.
- Publish your own ACP agent to reach every protocol-aware client.
- Evaluate agents side-by-side using pinned version numbers from the listings.
- Contribute new authentication-supporting agents to the open-source registry.
- Fetch the registry index programmatically to power agent discovery in a custom client.
Limitations
- A curated catalog/directory with no underlying AI model of its own; it lists ACP-compatible agents built by other vendors, so AI capability depends entirely on the individual agent you install.
- Listings are gated on authentication support and are maintainer-curated, so details like version numbers and descriptions are only as accurate as the submissions.
- It functions as a JSON index/CDN endpoint rather than a hosted agent runtime.
as of 2026-10-04
Verification history
We have re-verified Registry 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 Registry tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Free
$0/mo
Ideal for
Client developers, agent teams and open-source contributors who need read access to the ACP agent catalog and a submission path for their own agent.
What this tier adds
Free entry point: full read access to the catalog, programmatic fetch of registry.json via CDN, and fork-and-PR submission, with no account required.
Where the pricing makes sense
The company stage and team size where Registry's pricing actually pencils out — and where peers do it cheaper.
The registry itself is a free, read-only JSON catalog fetched from the CDN — no account needed to browse or pull it. The cost question sits downstream: each listed agent carries its own vendor pricing, from open-source CLI agents such as goose, Kilo and OpenCode that you self-fund, to commercial agents like Cursor, Devin and GitHub Copilot with their own seat and usage billing.
Setup time & first value
How long it actually takes to get something useful out of Registry — broken out by persona, not the marketing-page minute.
For a client developer, first value is a single curl plus JSON parsing — minutes if you already speak the protocol. For an agent team publishing a listing, budget an afternoon for the fork, the agent.json written against the schema, an optional 16x16 icon.svg, and the pull request, plus whatever review turnaround the maintainers take.
Switching to or from Registry
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From manual README tracking: replace per-vendor install docs with one registry.json fetch from the CDN endpoint.
- →From a hand-maintained agent list in your client: point at the versioned registry metadata and delete the list you were curating.
- →From vendor-specific installers: use the distribution info in each listing to drive automatic installation.
- ↗To a commercial agent marketplace: move if you need ratings, reviews, benchmark scores or a support contract the registry does not provide.
- ↗To vendor-direct distribution: if you only ever ship one agent, installing straight from that vendor's docs may be simpler than going through the catalog.
Integrations
Resources & Guides
- Resourceagentclientprotocol.com
Registry · Registry
Helpful link from agentclientprotocol.com
- Resourceagentclientprotocol.com
Introduction · Registry
Helpful link from agentclientprotocol.com
- Resourceagentclientprotocol.com
Architecture · Registry
Helpful link from agentclientprotocol.com
- Resourceagentclientprotocol.com
Overview · Registry
Helpful link from agentclientprotocol.com
Tutorials & Learning
YouTube returned 6 videos for “Registry”, and we withheld 6: 6 could not be judged, because “Registry” 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 Registry.
Official links
Featured Head-to-Head Comparisons
Registry vs Spider Cloud
If you need to discover ACP-compatible agents and contribute to a secure, open ecosystem, Registry is your tool — it's free and protocol-first. If you need real-time web data for AI agents or RAG pipelines, Spider Cloud offers a powerful, pay-as-you-go scraping engine with a 99.9% success rate and extensive integrations. Choose Registry for agent discovery; choose Spider Cloud for web data extraction.
Registry vs Presto Voice
Registry and Presto Voice serve completely different markets. Pick Registry if you're a developer needing a free, open-source directory of ACP-compatible agents with authentication. Pick Presto Voice if you're a QSR chain looking to automate drive-thru orders with voice AI and boost revenue. They aren't competitors.
Registry vs Temporal Ai
If you need to discover and distribute ACP-compatible agents with authentication, pick Registry. If you need to build reliable, crash-proof AI workflows with state persistence, retries, and human-in-the-loop, pick Temporal AI. They solve different problems—Registry is a catalog, Temporal is a runtime.
Popular in MCP Servers & Agent Tooling
Smithery
Smithery is an MCP server marketplace and connection layer that handles OAuth, credentials, and session state for your AI agent.
Arcade AI
Arcade AI is the MCP authorization runtime that enforces per-user permissions and logs every action your AI agents take.
Chrome DevTools MCP
Open-source MCP server that gives coding agents live Chrome DevTools for debugging, traces, and browser automation.
Frequently Asked Questions
Categories
Best-of guides
Topics
Used Registry? Help shape our editorial sentiment research.