--- title: "Connect your AI" url: "https://ailawframework.com/connect" type: "connect" site: "AI Law Framework" --- # Connect your AI > Connect Claude, ChatGPT, Claude Code or Codex to the AI Law Framework. Your AI asks before it assumes, researches the rules with sources, and keeps to a plan you can check. Canonical page: https://ailawframework.com/connect Site index for agents: https://ailawframework.com/llms.txt ## Bring the framework into Claude or ChatGPT Connect once, then ask for legal work the way you already do. Your AI works it through the framework: it asks before it assumes, checks the rules with sources, and keeps to a plan you can see — then hands you a record that is ready for your review. ### Your firm’s knowledge — sign-in Your AI also reads your firm’s approved AI policies and knowledge, as you. [Log in to connect your firm](https://ailawframework.com/sign-in?next=%2Fconnect) ### The public framework — no sign-in The framework, the practice library and precedent summaries. Nothing about you is kept. [Connect the public framework](https://ailawframework.com/connect#claude) [Add to Claude (opens claude.ai in a new tab)](https://claude.ai/customize/connectors?modal=add-custom-connector&connectorName=AI+Law+Framework&connectorUrl=https%3A%2F%2Failawframework.com%2Fmcp) Claude opens with the connector filled in. Select **Add** — it is free, with no sign-in. [ChatGPT](https://ailawframework.com/connect#chatgpt) · [Claude Code](https://ailawframework.com/connect#claude-code) · [Codex](https://ailawframework.com/connect#codex) On a Team or Enterprise plan, an Owner [adds it for the whole organisation](https://claude.ai/admin-settings/connectors?modal=add-custom-connector&connectorName=AI+Law+Framework&connectorUrl=https%3A%2F%2Failawframework.com%2Fmcp). Using ChatGPT, Claude Code or Codex instead? Each has its own steps below. ### How it works You ask as usual — “draft the letter of advice”, “build the chronology”, “check this lease”. Behind the scenes your AI calls the framework’s tools one step at a time, and each step only opens once the one before it is done. What you get from each: 1. #### The right questions, once Define the task It asks what a careful supervising lawyer would — what you are producing, who relies on it, the forum, any rule on AI use — in one short message, and picks up whatever your request already said. 2. #### Sources you can check Assess and decide It researches the court’s or regulator’s rules on AI, the law the work relies on, the framework’s guidance for the task and its own product’s terms, and quotes each source with the date it read it. 3. #### A straight answer on whether to use AI Plan the human work It scores itself on the framework’s three questions against fixed wording, with doing the work by hand scored beside it, so you see when AI helps and when it does not. 4. #### A plan everyone can follow Do the work Each stage gets its level — who drafts, who checks what — and the AI keeps to it. At Red it writes nothing into the document; sworn evidence is always Red. 5. #### A record ready for the file Check, approve and record You receive a work record listing every high-risk point against its source, with the checks, the approval and the release left for you. Every answer built this way names the AI Law Framework and links here, so a partner, a client or a court can see which method was applied. The AI proposes; the responsible lawyer decides. ### Connect your firm’s knowledge There are two addresses. Use one of them, not both: the firm connector already includes everything the public one has. | Address | Sign-in | What your AI can read | | --- | --- | --- | | `https://ailawframework.com/mcp` | None. Nothing about you or your task is kept. | The framework, its practice library, and precedent summaries with their AI-use guidance. | | `https://ailawframework.com/mcp/firm` | Log in on this site and approve the app once. | All of the above, plus your firm’s approved AI policies, its knowledge base and the full text of the precedent library, as you. It can save drafts only if you allow it. | In Claude, add a custom connector with the firm address and select **Connect**; in Claude Code, run the command below. Your AI sends you here to log in and shows you exactly what it is asking for before you approve. Firm connector address: ``` https://ailawframework.com/mcp/firm ``` Claude Code command for the firm connector: ``` claude mcp add --transport http ailawframework-firm https://ailawframework.com/mcp/firm ``` What is kept for the firm connector: which app you approved and when, its keys in a one-way hashed form, and an access log your firm’s owners and administrators can read (which app read which item, and when; never your questions or any text). Steps for ChatGPT, Codex and Cursor, your connected apps and the Disconnect button are at [Firm › AI agents](https://ailawframework.com/firm/agents). ### Claude: web, desktop and mobile Select **Add to Claude**: Claude opens its Add custom connector dialog with the name and address filled in, and you confirm. Or open **Customize › Connectors**, select **+ Add**, then **Add custom connector**, name it AI Law Framework, paste this address, select **Continue** and choose **No sign in**. Once added on claude.ai it is there in the desktop and mobile apps too. [Add to Claude (opens claude.ai in a new tab)](https://claude.ai/customize/connectors?modal=add-custom-connector&connectorName=AI+Law+Framework&connectorUrl=https%3A%2F%2Failawframework.com%2Fmcp) Claude opens with the connector filled in. Select **Add** — it is free, with no sign-in. MCP server address: ``` https://ailawframework.com/mcp ``` The framework’s prompts are in the connector’s menu in a conversation, ready to start a task with: - **Apply the AI Law Framework to a legal task** - **Get started with the AI Law Framework** - **Define the task: the clarifying questions** - **Research the evidence before scoring** - **Assess and decide on the three pillars** - **Plan the human work, then do only that** - **Hand over for checks, approval and release** Plans and steps, checked on 18 October 2026: [Claude Help Center, custom connectors](https://support.claude.com/en/articles/11175166-get-started-with-custom-connectors-using-remote-mcp). ### ChatGPT ChatGPT adds the framework as a plugin from its server address. You do this on ChatGPT on the web: OpenAI does not offer it in the mobile app, and there is no developer mode to turn on first. ChatGPT has no link that fills its form in for you, so copy the address and enter the values below. MCP server address: ``` https://ailawframework.com/mcp ``` #### On the web 1. Open [Plugins](https://chatgpt.com/plugins) in ChatGPT. 2. Select the **+** button, then **Add custom MCP server**. 3. Enter the name, paste the address under **Connection**, and set authentication as shown. 4. Tick **I understand and want to continue**, then select **Create as a plugin** and check the tools it found. 5. Install the plugin. In a new chat, type **@**, choose AI Law Framework and ask for the work. #### What to enter | Name | AI Law Framework | | --- | --- | | Connection (MCP server URL) | `https://ailawframework.com/mcp` | | Authentication | None: the public server has no sign-in | Once you have chosen the plugin with @, ask for the work the way you usually would. For example: A first request to try: ``` Draft a letter of advice to my client about a residential lease dispute in NSW. ``` #### Who can add it | ChatGPT plan | What OpenAI’s help article says | | --- | --- | | Business | Only admins and owners can add it and publish it to the workspace. Members then install it from the workspace’s plugins. | | Enterprise and Edu | Admins and owners, and members an admin has given the Create plugins with MCPs permission. Only admins and owners publish it to the workspace. | | Pro | OpenAI says Pro users can connect custom MCP servers with read and fetch permissions. Every tool on the public server only reads, so all of it works on Pro. | | Plus and Free | OpenAI’s article does not say whether these plans can add a custom MCP server. | For your firm’s knowledge, use `https://ailawframework.com/mcp/firm` instead and choose OAuth as the authentication: ChatGPT sends you here to log in and approve. On Pro, which is limited to read and fetch, the firm connector reads but cannot save drafts. #### In the ChatGPT desktop app 1. Open **Settings › MCP servers** and select **Add server**. 2. Name it ailawframework, choose **Streamable HTTP** and paste the address above. 3. Save, then select **Restart**. The desktop app shares this with Codex, so it is set up there as well. One-click install comes later. Once the AI Law Framework is listed in ChatGPT’s plugin directory, its listing will have an **Install** button and this page will link to it in place of these steps. Checked against OpenAI’s own pages on 18 October 2026: [Connecting custom MCP servers in ChatGPT](https://help.openai.com/en/articles/12584461-connecting-custom-mcp-servers-in-chatgpt) (web steps and plans) and [ChatGPT docs, MCP](https://learn.chatgpt.com/docs/extend/mcp) (desktop app). ### Claude Code Run this once in a terminal and it is available in every project on this computer: Claude Code command: ``` claude mcp add --transport http --scope user ailawframework https://ailawframework.com/mcp ``` The prompts arrive as slash commands — start with `/mcp__ailawframework__ai-law-framework`. ### Codex Run this once in a terminal: Codex command: ``` codex mcp add ailawframework --url https://ailawframework.com/mcp ``` Or add it to `~/.codex/config.toml` yourself: Codex configuration: ``` [mcp_servers.ailawframework] url = "https://ailawframework.com/mcp" ``` ### Any other app that takes an MCP server It is a standard remote MCP server over Streamable HTTP, with no sign-in. Give your app the address above, or this configuration: MCP configuration: ``` { "mcpServers": { "ailawframework": { "type": "http", "url": "https://ailawframework.com/mcp" } } } ``` ### Questions people ask **What does connecting it actually change?** Your AI gains the framework’s tools and its practice library. When you ask it for a piece of legal work, it works through the same five steps this site does: it asks what it needs to know, researches the rules and the law with sources, scores whether it should be used for this task at all, keeps to what the plan allows at each stage, and gives you a work record to check, approve and file. **Do I need an account here, or to pay?** Not for the framework. The public connector (/mcp) is free and needs no sign-in on this site. To give your AI your own firm’s policies, knowledge and the full precedent texts, add the signed-in connector (/mcp/firm) instead: it asks you to log in here once. Your AI plan decides whether you can add a connector at all. Claude allows custom connectors on Free (one connector), Pro, Max, Team and Enterprise plans. In ChatGPT, admins and owners of Business, Enterprise and Edu workspaces add a custom MCP server from Plugins (on Enterprise and Edu, so can members an admin permits), and Pro users can connect one with read and fetch permissions, which covers everything on /mcp. OpenAI’s article does not say whether Plus or Free can. **What is the difference between /mcp and /mcp/firm?** Both carry the framework’s tools and its practice library. /mcp is public: no sign-in, nothing about you is known, and precedents show their summary and AI-use guidance only. /mcp/firm needs you to log in on this site and approve the app once; it then also reads your firm’s approved AI policies, its knowledge base and the full text of the precedent library, as you, and can save drafts if you allow it. Use one or the other, not both. **Claude shows a warning when I add it. Is that normal?** Yes. Claude shows a caution whenever a connector is suggested by a link on a website, so you can check that the name and address come from a developer you trust: they should read AI Law Framework and https://ailawframework.com/mcp. It also notes that a connector without sign-in can be used by anyone who has its address. That is by design for /mcp: it is free, keeps nothing about your task and returns only the framework’s own published material. Leave Authentication on No sign in, add no headers, and select Add. The firm connector (/mcp/firm) is different: Claude sends you to this site to log in and approve it. **Is there a one-click Add to ChatGPT button?** Not yet. ChatGPT has no link that fills in its Add custom MCP server form, so this page gives you the address to copy and the exact values for each field. A one-click Install button comes once the AI Law Framework is listed in ChatGPT’s plugin directory; this page will link to it then. Claude’s Add to Claude button does fill in its form, and you still confirm before anything is added. **We are on a Team or Enterprise plan. Who adds it?** In Claude, an Owner adds it once for the organisation, and each member then connects to it and switches it on. In ChatGPT Business, an admin or owner adds it from Plugins and publishes it to the workspace, and members install it from there; on Enterprise and Edu, members an admin has given the Create plugins with MCPs permission can add it too, but only admins and owners publish it. The organisation link on this page opens Claude’s dialog for an Owner. **Will it slow me down?** Only by the minute it takes to answer the questions a careful supervising lawyer would ask anyway — and it picks up whatever you already said in your request. “Not sure” is always an acceptable answer; it simply raises the level of human involvement by one step, which is what doubt should do. **Can the AI skip a step or change an answer later?** No. Each step needs the signed record the previous step returned, and the server refuses a record that has been altered. An answer the AI assumed instead of asking is refused too. What the server cannot do is check that the AI reported your words faithfully, so the work record shows where every answer and finding came from, for the lawyer to see. **What does the server see, and what does it keep?** It sees what your AI sends to its tools: the task as described, your answers and the findings with their sources. It uses them to apply the rules and keeps nothing about the task — no database row, no copy of the record — on either connector. Describe the task without client names where you can; the framework never needs them. The public connector (/mcp) keeps nothing at all. The firm connector (/mcp/firm) also keeps what a sign-in needs: which app you approved and when (your connection), the connection’s keys in a form that cannot be used if copied (only a one-way hash), and, for your firm’s owners and administrators, an access log of which app read which of your firm’s items and when. The log holds item ids, never your questions or any text. **Who makes the decision?** The responsible lawyer. The AI proposes its own scores against the framework’s fixed wording, the evidence it found caps the governance score, and the record says plainly that they are a proposal for the lawyer to confirm or change. **What happens when the answer is “do not use”?** The AI tells you so, with the reasons, and does not produce the output. Doing the work by hand is always one of the options and is scored beside the AI, so you still leave with a plan and a record — and often the reason is something you can fix, such as confirming a supplier term. **Does it work outside Australia?** Yes. The method is the same anywhere: the AI researches the rules of the court, tribunal or regulator that will receive the work. The library is richest for Australian jurisdictions and also carries New Zealand and United States versions of many tasks. **Is it legal advice, or a certificate that the work is right?** Neither. It is a method for deciding how AI may be used on a task and for recording the human checks. It does not certify that a tool is safe, that the work is correct or that a saving is real; every judgement in the record is one a person makes or confirms. **Why do answers mention the AI Law Framework?** So whoever reads them — a partner, a client, a court — knows which method was applied and can look it up. Every result the tools return carries the name and a link to this site, and the work record ends with it. **Who at my firm can see what the firm connector reads?** Each connection reads as the person who approved it, so it sees exactly what that person may see. Your firm’s owners and administrators can see the access log (who, which app, which tool, which items, when) and can switch AI agents off for the whole firm, which disconnects every app at once. Anyone can disconnect their own apps at Firm › AI agents. **How do I remove it?** In Claude, remove the connector at Customize › Connectors; in ChatGPT, remove the plugin from Plugins, where you added it. In a terminal, run “claude mcp remove ailawframework” or “codex mcp remove ailawframework” (or ailawframework-firm). For the firm connector, also select Disconnect at Firm › AI agents: that cancels its access at once, even on a device you no longer have. [Read the method](https://ailawframework.com/how-it-works.md) · [See it applied to one task](https://ailawframework.com/how-it-works/example) · [Privacy](https://ailawframework.com/privacy.md) · [Ask us something else](https://ailawframework.com/enquiries)