copilotkit 4.18.0 → 4.20.0

package/onboarding/index.json

@@ -1,5 +1,5 @@
 {
-  "graphTree": "07aa5486282bd324b41fecac4685d20b5ed11bfa",
+  "graphTree": "da1c6185427d9e627b0d6a6b6abda10dea05e09a",
   "intentRoots": {
     "add-a2ui": "feature/a2ui/start",
     "add-chat-suggestions": "feature/chat-suggestions/start",

package/onboarding/prompts/authenticate/start.md

@@ -13,23 +13,36 @@ does is hidden from the developer.
 ## Command approvals
 
 Run each CLI call exactly as written, as its own command: no pipe, `tee`, `&&`, redirect,
-or variable prefix. If your harness asks the developer to approve commands, request
-escalation on the `onboard identify` call with the prefix rule
-`["npx", "--prefer-offline", "--yes", "copilotkit@4.18.0"]`. Every CLI call in this run
-matches it. Say then that the install, the dev servers, `curl` to the local app, and `ps`,
-`lsof`, and `kill` ask for approval later.
+or variable prefix. Every CLI call in this run matches the prefix rule
+`["npx", "--prefer-offline", "--yes", "copilotkit@4.20.0"]`. If a command can request
+escalation with a prefix rule, request this one on the `onboard identify` call. If your
+harness asks the developer to approve commands, tell them one approval of that prefix covers
+every CLI call. Say then that the install, the dev servers, `curl` to the local app, and
+`ps`, `lsof`, and `kill` ask for approval later.
+
+Never add or edit a permission rule or a settings file yourself. That choice belongs to
+the developer.
+
+If your harness refuses a CLI call without asking the developer, do not retry it or run it
+another way. Name the refused command, ask the developer to approve it or to switch to a
+mode that asks them, and end your turn. When they answer, run it once more.
+That refusal is a pause, not a stop.
 
 ## First, say which coding agent you are
 
 Run this before anything else in this prompt, including the welcome message below:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard identify --coding-agent <coding-agent-slug> --model <model-id>
+npx --prefer-offline --yes copilotkit@4.20.0 onboard identify --coding-agent <coding-agent-slug> --model <model-id> --model-family <model-family>
 ```
 
-Report your own product with one short slug, such as `codex` or `claude-code`. For
-`--model`, give the exact model id your instructions name. If they name none, leave out
-`--model` rather than guess.
+Report your own product with one short slug, such as `codex` or `claude-code`.
+
+- `--model`: copy the model id exactly as your instructions write it. If they name no
+  exact model id, leave out `--model` rather than guess.
+- `--model-family`: your model's family, such as `claude-opus` or `gpt-5`. Give it even
+  without `--model`, and leave out `--model-family` if you do not know it.
+- Never give your product slug as a model or a family.
 
 ## How to run this onboarding
 
@@ -49,10 +62,10 @@ printed once.
 Every later command must run from the directory this run started in. A command run
 somewhere else is refused, because the run it belongs to cannot be found from there. That
 refusal is a step, not a stop to report: run
-`npx --prefer-offline --yes copilotkit@4.18.0 onboard start --run <onboarding_run_id>` where you are
+`npx --prefer-offline --yes copilotkit@4.20.0 onboard start --run <onboarding_run_id>` where you are
 standing, then run the refused command again.
 
-If you no longer hold the id, run `npx --prefer-offline --yes copilotkit@4.18.0 onboard start` in the
+If you no longer hold the id, run `npx --prefer-offline --yes copilotkit@4.20.0 onboard start` in the
 directory this run started in, with no `--run` and no `--intent`. It reprints the id of the
 run under way and starts nothing. An id this project does not already hold starts a second
 run, and the work already done stops counting toward the first.
@@ -119,7 +132,7 @@ the run waits is a pause, not a stop, even when your harness ends it. A terminal
 the report itself, or says to file none, overrides this.
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard friction --phase stop --category <slug> --message "<sentences>"
+npx --prefer-offline --yes copilotkit@4.20.0 onboard friction --phase stop --category <slug> --message "<sentences>"
 ```
 
 `--message` takes one or two sentences: the step you stopped at and what stopped it.
@@ -134,7 +147,7 @@ rule that stopped you says next: send the report, then stop.
 
 ## Check sign-in
 
-Run `npx --prefer-offline --yes copilotkit@4.18.0 login --json` before any project work. This command
+Run `npx --prefer-offline --yes copilotkit@4.20.0 login --json` before any project work. This command
 checks the current session. It starts sign-in only as needed. Treat this as a long-lived
 streaming process. Do not wait for the command to exit before you read its standard output.
 
@@ -147,42 +160,36 @@ Read each JSON Lines record as the running command writes it:
    When that first URL arrives, send this exact message once. Do not send it when the
    existing session is already valid, and do not add a feature list or another explanation:
 
-   I will open a sign-in page. Signing in gives this project a managed CopilotKit project
+   I am opening a sign-in page. Signing in gives this project a managed CopilotKit project
    and its API key. That key is what lets your users reopen past conversations. You can
    remove it later to disconnect this workspace.
 
    If it carries `user_code`, add: `Check that the page shows this code: <user_code>`.
 
-3. Open the first `authentication_url` exactly once with the operating system's default
-   browser. Your own built-in browser is not that opener and does not carry the session
-   the developer already signed in with. Run the opener separately from the long-lived
-   login process. Use the command for the host operating system:
-   - macOS: `open "<authentication_url>"`
-   - Linux: `xdg-open "<authentication_url>"`
-   - Windows PowerShell: `Start-Process "<authentication_url>"`
-     Pass the exact URL as one quoted argument, unchanged.
-4. If the matching opener is unavailable or exits with an error, show the clickable URL to
-   the developer and ask them to finish sign-in there. Do not retry the opener.
-5. Keep monitoring the original login process after the open attempt. Do not start a
-   second login command. Do not treat the opener command finishing as sign-in finishing.
-6. Continue onboarding only after that same process emits `type: completed`.
-7. On `type: failed` with `reason: expired`, ask once whether to retry. If yes, start one
+3. The CLI opens the page in the default browser itself when it can, and the
+   `action_required` record says whether it did. With `"browser_opened": true`, do not open
+   it again. With `"browser_opened": false`, show the clickable URL to the developer and ask
+   them to finish sign-in there. The page works from any device. Your own built-in browser
+   does not carry the session the developer already signed in with, so do not use it.
+4. Keep monitoring the same login process. Do not start a second login command.
+5. Continue onboarding only after that same process emits `type: completed`.
+6. On `type: failed` with `reason: expired`, ask once whether to retry. If yes, start one
    new `login --json`. Use `--force` only when the developer asks.
-8. If that process emits `type: failed`, report the error and stop onboarding. On
+7. If that process emits `type: failed`, report the error and stop onboarding. On
    `reason: repeated_expiry`, rerun `login` with `--force` only if the developer asks.
 
 If your harness ends the login process or your turn before the process emits
 `type: completed` or `type: failed`, that is a pause, not a failed sign-in. If the process
 still runs, end your turn and wait for the developer to say they signed in, then read the
 same process. If the process ended without emitting `type: failed`, run `login --json`
-again and handle a new `authentication_url` as above. Step 5 forbids a second login only while the first one runs.
+again and handle a new `authentication_url` as above. Step 4 forbids a second login only while the first one runs.
 
 Do not use `whoami` for this check. Do not put account tokens or credentials in the
 conversation.
 
 When sign-in is settled, run
-`npx --prefer-offline --yes copilotkit@4.18.0 onboard read research/gather`.
+`npx --prefer-offline --yes copilotkit@4.20.0 onboard read research/gather`.
 
 If sign-in stops onboarding, run
-`npx --prefer-offline --yes copilotkit@4.18.0 onboard read stopped/run-failed`. It says nothing about
+`npx --prefer-offline --yes copilotkit@4.20.0 onboard read stopped/run-failed`. It says nothing about
 whether this project's stack is supported, which is not yet known at this point.

package/onboarding/prompts/conversion/plan.md

@@ -100,7 +100,7 @@ of this journey, and a drawer that lists nothing is not a reason to mint one. Li
 tokens stay a self-hosted concern.
 
 Where this journey's frontend framework ships no threads drawer -- React Native --, the
-fifth item is proved by `npx --prefer-offline --yes copilotkit@4.18.0 verify --round-trip`, which sends a
+fifth item is proved by `npx --prefer-offline --yes copilotkit@4.20.0 verify --round-trip`, which sends a
 real request and reads the answer back off the thread it landed on. The run says which of
 the two it proved. Do not send a run to the managed Intelligence dashboard to prove this.
 An authenticated browser session at a hosted dashboard is the one thing an agent-driven run
@@ -116,7 +116,7 @@ and passed -- so a run scored on either one was scored on the same fact. Every r
 resolves that condition from, while `licenseStatus` is a compatibility projection of it
 for older thread surfaces and outlives them only until those surfaces go.
 
-Then run `npx --prefer-offline --yes copilotkit@4.18.0 onboard read credentials/finalize-plan`.
+Then run `npx --prefer-offline --yes copilotkit@4.20.0 onboard read credentials/finalize-plan`.
 
 If a selected page does not load after the second method, run
-`npx --prefer-offline --yes copilotkit@4.18.0 onboard read unsupported/no-validated-path`.
+`npx --prefer-offline --yes copilotkit@4.20.0 onboard read unsupported/no-validated-path`.

package/onboarding/prompts/credentials/finalize-plan.md

@@ -24,7 +24,7 @@ Choose the starting state from the repository findings:
 A run served `proof/oss-baseline` records one of the last two.
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard classify --starting-state <starting-state> --agent-framework <agent-framework> --frontend <frontend>
+npx --prefer-offline --yes copilotkit@4.20.0 onboard classify --starting-state <starting-state> --agent-framework <agent-framework> --frontend <frontend>
 ```
 
 If a value is refused, fix it from the choices the earlier prompts gave, and run the
@@ -37,7 +37,7 @@ runtime `.env` file, that directory is the target, not the repository root. Do n
 the developer to another terminal.
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard credentials --json
+npx --prefer-offline --yes copilotkit@4.20.0 onboard credentials --json
 ```
 
 Every credential this run writes goes to the target's `.env`, the environment path. The
@@ -67,9 +67,9 @@ project from a listing without an answer that names it.
 `--runtime-url` is the URL port research settled, mount path included:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard credentials --create <name> \
+npx --prefer-offline --yes copilotkit@4.20.0 onboard credentials --create <name> \
   --runtime-url http://localhost:<port>/api/copilotkit --json
-npx --prefer-offline --yes copilotkit@4.18.0 onboard credentials --project <slug-or-id> \
+npx --prefer-offline --yes copilotkit@4.20.0 onboard credentials --project <slug-or-id> \
   --runtime-url http://localhost:<port>/api/copilotkit --json
 ```
 
@@ -79,8 +79,8 @@ command unless the developer asks to see their projects, or `notFound` is presen
 show only what they asked for, and do not order it by creation time:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 project list --json
-npx --prefer-offline --yes copilotkit@4.18.0 project list --search <query> --json
+npx --prefer-offline --yes copilotkit@4.20.0 project list --json
+npx --prefer-offline --yes copilotkit@4.20.0 project list --search <query> --json
 ```
 
 `app-directory` means that the key landed where no app reads it. Pick the directory in
@@ -95,4 +95,4 @@ the `--project` command from there with `project.slug`. Do not copy or link the
   reused. The round-trip proof later runs the authenticated Intelligence checks, so record
   the reused project credentials as unverified in the plan until those checks pass.
 
-Then run `npx --prefer-offline --yes copilotkit@4.18.0 onboard read credentials/settle-credentials`.
+Then run `npx --prefer-offline --yes copilotkit@4.20.0 onboard read credentials/settle-credentials`.

package/onboarding/prompts/credentials/plan.md

@@ -20,7 +20,33 @@ If the target directory holds no project and the copied prompt came from the Sla
 Microsoft Teams docs page, do not ask the framework question. The Channel starter brings
 its own agent, so an answer here decides nothing. Tell the developer that the Channel
 starter decides the agent, then run
-`npx --prefer-offline --yes copilotkit@4.18.0 onboard read frontend/plan`.
+`npx --prefer-offline --yes copilotkit@4.20.0 onboard read frontend/plan`.
+
+## Frameworks this machine cannot run yet
+
+The research findings carry the CLI's toolchain reading. Its `unready` list names each
+framework that this machine cannot run until a toolchain is installed, and the toolchain
+it needs. Read that list before you ask the framework question, so the developer learns
+about a missing toolchain before they choose, and not after.
+
+- Do not recommend a framework on that list.
+- When you show a framework on that list, show it after the ready ones. In the same line,
+  name the toolchain it needs and its install page.
+- Do not install a toolchain yourself.
+
+If the developer chooses a framework on that list, or the repository already has an agent
+in one, keep that framework. Before the run changes the project, tell the developer three
+things in one message:
+
+1. Which toolchain to install, and its install page.
+2. That the coding agent's shell reads PATH when it starts, so it cannot see the new
+   toolchain until the developer restarts the coding agent.
+3. That the run resumes after the restart with
+   `npx --prefer-offline --yes copilotkit@4.20.0 onboard start --run <onboarding_run_id>`.
+
+Then end your turn and wait for the developer. That restart is a step here, not an error.
+
+## Choose the framework
 
 Use a gentle wizard. Ask one short question at a time. Do not combine separate choices in
 one question. Put the recommendation first, give its evidence in one sentence, and keep
@@ -75,26 +101,26 @@ developer's words instead.
 
 Use exactly one matching internal route:
 
-1. AG2: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/ag2`
-2. Agno: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/agno`
-3. Built-in CopilotKit agent: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/built-in`
-4. Claude Agent SDK Python: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/claude-sdk-python`
-5. Claude Agent SDK TypeScript: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/claude-sdk-typescript`
-6. CrewAI Flows: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/crewai-flows`
-7. Deep Agents: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/deep-agents`
-8. LangGraph Python: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/langgraph-python`
-9. LangGraph FastAPI: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/langgraph-fastapi`
-10. LangGraph TypeScript: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/langgraph-typescript`
-11. LlamaIndex: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/llamaindex`
-12. ADK: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/google-adk`
-13. Microsoft Agent Framework Python: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/ms-agent-python`
-14. Microsoft Agent Framework .NET: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/ms-agent-dotnet`
-15. Mastra: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/mastra`
-16. MS Agent Harness .NET: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/ms-agent-harness-dotnet`
-17. Pydantic AI: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/pydantic-ai`
-18. Strands Agents Python: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/strands-python`
-19. Strands Agents TypeScript: `npx --prefer-offline --yes copilotkit@4.18.0 onboard read framework/strands-typescript`
+1. AG2: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/ag2`
+2. Agno: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/agno`
+3. Built-in CopilotKit agent: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/built-in`
+4. Claude Agent SDK Python: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/claude-sdk-python`
+5. Claude Agent SDK TypeScript: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/claude-sdk-typescript`
+6. CrewAI Flows: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/crewai-flows`
+7. Deep Agents: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/deep-agents`
+8. LangGraph Python: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/langgraph-python`
+9. LangGraph FastAPI: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/langgraph-fastapi`
+10. LangGraph TypeScript: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/langgraph-typescript`
+11. LlamaIndex: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/llamaindex`
+12. ADK: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/google-adk`
+13. Microsoft Agent Framework Python: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/ms-agent-python`
+14. Microsoft Agent Framework .NET: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/ms-agent-dotnet`
+15. Mastra: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/mastra`
+16. MS Agent Harness .NET: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/ms-agent-harness-dotnet`
+17. Pydantic AI: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/pydantic-ai`
+18. Strands Agents Python: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/strands-python`
+19. Strands Agents TypeScript: `npx --prefer-offline --yes copilotkit@4.20.0 onboard read framework/strands-typescript`
 
 If the project has an agent in another framework, or no listed framework fits, keep the
 developer's current agent and run
-`npx --prefer-offline --yes copilotkit@4.18.0 onboard read unsupported/no-validated-path`.
+`npx --prefer-offline --yes copilotkit@4.20.0 onboard read unsupported/no-validated-path`.

package/onboarding/prompts/credentials/settle-credentials.md

@@ -21,7 +21,7 @@ The plan names the model credential variables. Run this from the directory that
 `environmentPath` in the project step's result, with one `--model-key` for each variable:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard credentials --model-key <variable> --json
+npx --prefer-offline --yes copilotkit@4.20.0 onboard credentials --model-key <variable> --json
 ```
 
 When the framework node names another env file for the credential, add
@@ -31,25 +31,39 @@ never prints a value. When every key settles, it protects the credential paths.
 `status`:
 
 - `ready`: continue. Name each `unsupported` or `undetermined` `reason` in `modelKeys` in
-  the closing report.
-- `needs-developer` with `question.status` `missing`: ask the developer where the
-  credential lives.
+  the closing report. When `outsideRepository` names a file, the app does not load it by
+  itself. By default, plan to copy the key into the app's env file, which git ignores.
+  This command protects that env file, and the copy is a file the run writes itself, so
+  list the copy under `Files the run writes itself`, not under `Authorization requested`.
+  The developer's approval of the plan is their consent to the copy. Copy the key without
+  printing it. Then record the copy with
+  `npx --prefer-offline --yes copilotkit@4.20.0 onboard protect --accept-credential --path <env file>`. If the developer declines the copy, the plan has the app read the key from an
+  environment variable that the developer sets. The plan then says how to set it. Never
+  write the outside path into a project file, not even in a config file such as
+  `langgraph.json`. A committed file that names a folder in the developer's home directory
+  gives every other machine no key, and `onboard audit` fails on a changed file that names
+  it. In the closing report, say where the app reads the key from.
+- `needs-developer` with `question.status` `missing`: ask the developer `question.ask` in
+  those words. It offers three answers: another env file in this project, a file outside it,
+  or the developer adds the key to `question.envFile`. If they name a file, rerun the command
+  with `--env-file <path>`. If they add the key, rerun it as it is. Do not replace the
+  question with an instruction to paste the value into a file. The key can already be in a
+  file that you can point the check at.
 - `needs-developer` with `question.status` `fail`: tell the developer `question.cause`.
   `model_quota` means that the key has no credits, and `model_auth` means that the vendor
   rejected it. Offer three choices: add credits, use another key, or switch the model
   provider.
 - `stopped`: the command already filed the stop report. Stop onboarding.
 
-Until the developer answers, the run waits. When you ask, name each variable and the file
-it goes in, and say that the run continues when they reply or resume this session. Under a
-harness that ends the session with your turn, such as `codex exec`, that message is the only
-thing the developer sees. This is a pause, not a stop. The command already reported it.
+Until the developer answers, the run waits. When you ask, name each variable, and say that
+the run continues when they reply or resume this session. Under a harness that ends the
+session with your turn, such as `codex exec`, that message is the only thing the developer
+sees. This is a pause, not a stop. The command already reported it.
 End your turn. When the developer answers, run the command again. Do not write a
 placeholder or an empty value.
 
 Never read a file outside the project directory to find a credential: a key found that way
-bills another project. Asking for a path is not requesting a secret, and neither is asking
-the developer to write the value into the env file. A path the developer names is theirs
-to give, including one outside the project. A path this run finds is not.
+bills another project. Asking for a path is not requesting a secret. A path the developer
+names is theirs to give, including one outside the project. A path this run finds is not.
 
-Then run `npx --prefer-offline --yes copilotkit@4.18.0 onboard read credentials/write-plan`.
+Then run `npx --prefer-offline --yes copilotkit@4.20.0 onboard read credentials/write-plan`.

package/onboarding/prompts/credentials/write-plan.md

@@ -11,7 +11,7 @@ friction command follows the telemetry setting the developer already chose, so i
 separate question.
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard friction --phase stop --category <slug> --message "<sentences>"
+npx --prefer-offline --yes copilotkit@4.20.0 onboard friction --phase stop --category <slug> --message "<sentences>"
 ```
 
 `--message` takes one or two sentences: the step you stopped at and what stopped it.
@@ -72,7 +72,7 @@ page is selected yet, add the one for the frontend selected above to the same li
    Native, so the planner proves the thread with `verify --round-trip` instead.
 
 Spawn one planning subagent. Tell it to run
-`npx --prefer-offline --yes copilotkit@4.18.0 onboard read subagent/create-plan` first and follow the prompt
+`npx --prefer-offline --yes copilotkit@4.20.0 onboard read subagent/create-plan` first and follow the prompt
 it returns. If that read fails because the subagent cannot use the shell, stop that subagent.
 Run the same command yourself, then spawn a fresh subagent with the returned prompt and the
 same handoff. Give it the repository findings, selected framework, frontend, model, credential
@@ -83,7 +83,7 @@ Wait for the subagent to finish.
 Continue only if the planning result starts with `Status: passed`. For `Status: failed`,
 send the result back to the planning subagent for repair, up to three attempts. For
 `Status: blocked`, or a third failed result, run
-`npx --prefer-offline --yes copilotkit@4.18.0 onboard read stopped/run-failed`. A plan this run cannot
+`npx --prefer-offline --yes copilotkit@4.20.0 onboard read stopped/run-failed`. A plan this run cannot
 write is a run that broke, not a stack the documentation does not cover. The exception is
 a blocked result that names `unsupported/no-validated-path`: the planning subagent found
 that no page supports the plan, so take the no-supported-path route at the end of this
@@ -96,7 +96,7 @@ for, implementation steps, validation steps, and proof steps. Where the plan upg
 CopilotKit dependencies, it must name the versions it moves, every other pin it has to
 move with them, and the revert it falls back to. It must not contain secret values.
 
-Show only the selected framework, frontend, model, planned file changes with the feature explanations below where applicable, any ignore-file line this run adds, the CopilotKit dependency upgrade and the versions it moves, every other dependency the upgrade moves and the versions it moves them to, validation commands, and proof steps.
+Show only the selected framework, frontend, model, planned file changes with the feature explanations below where applicable, any ignore-file line this run adds, the files the run writes itself, the CopilotKit Skills install as a choice the developer approves or declines, the CopilotKit dependency upgrade and the versions it moves, every other dependency the upgrade moves and the versions it moves them to, validation commands, and proof steps.
 
 Explain these benefits in your own words within the planned changes, only when the
 plan includes the feature:
@@ -111,6 +111,9 @@ plan includes the feature:
 These explanations are required content, not fixed wording. They explain the existing
 plan. They do not add features, implementation steps, or a separate message or question.
 
+Carry the developer's answer on the Skills install into the implementation and proof
+handoffs. Approving the plan without a word on the Skills approves the install.
+
 An upgrade to a working install is the developer's call, so it belongs in the plan they
 approve rather than in the implementation that follows. Do not upgrade a dependency the
 developer did not approve.
@@ -131,7 +134,7 @@ build. Keep this window open until the app runs." Decide from your own approval
 as the welcome did, not from your coding-agent slug.
 
 If the developer approves the plan, run
-`npx --prefer-offline --yes copilotkit@4.18.0 onboard read implementation/build-and-validate`.
+`npx --prefer-offline --yes copilotkit@4.20.0 onboard read implementation/build-and-validate`.
 
 If no exact supported path or documentation URL exists, run
-`npx --prefer-offline --yes copilotkit@4.18.0 onboard read unsupported/no-validated-path`.
+`npx --prefer-offline --yes copilotkit@4.20.0 onboard read unsupported/no-validated-path`.

package/onboarding/prompts/fallback/best-effort.md

@@ -16,7 +16,7 @@ you stop. The friction command follows the telemetry setting the developer alrea
 so it needs no separate question.
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard friction --phase stop --category <slug> --message "<sentences>"
+npx --prefer-offline --yes copilotkit@4.20.0 onboard friction --phase stop --category <slug> --message "<sentences>"
 ```
 
 `--message` takes one or two sentences: the step you stopped at and what stopped it.
@@ -62,19 +62,19 @@ or frontend, stop onboarding.
 
 Use these rules for every protected-path check in this fallback:
 
-- Run `npx --prefer-offline --yes copilotkit@4.18.0 onboard audit` from the target app directory.
+- Run `npx --prefer-offline --yes copilotkit@4.20.0 onboard audit` from the target app directory.
 - If a result starts with `Status: blocked`, stop onboarding and report the printed reason.
   It proved nothing changed, so do not report a preservation failure.
 - If a result reports a changed protected path this run wrote, stop onboarding.
 - If a result reports a changed env file, this rule replaces the one above, even when this
   run wrote the file. This run asked the developer to place a
   credential there, so take the credential route rather than the external one. Accept it
-  with `npx --prefer-offline --yes copilotkit@4.18.0 onboard protect --accept-credential --path <path>`,
+  with `npx --prefer-offline --yes copilotkit@4.20.0 onboard protect --accept-credential --path <path>`,
   run the audit again, and name it in the closing report. If the command refuses because
   it names a lost variable, stop onboarding.
 - Before any proof subagent runs, a changed path that no Files changed section from this
   run names came from outside the run. Accept it by name with
-  `npx --prefer-offline --yes copilotkit@4.18.0 onboard protect --accept-external --path <path>`, run the
+  `npx --prefer-offline --yes copilotkit@4.20.0 onboard protect --accept-external --path <path>`, run the
   audit again, and name it in the closing report.
 - After a proof subagent runs, a changed path that no Files changed section from this run
   names is unsettled rather than outside the run, because the proof subagent here returns
@@ -95,8 +95,10 @@ files, configuration files, dependencies, or tracked files. Allow only operation
 to project-owned processes, ports, and request options.
 Require this result format: Start with `Status: passed`, `Status: failed`, or `Status: blocked`.
 Tell it to use `Status: passed` only when the proof attempt completed with `performed`,
-`skipped-no-browser-tool`, or `skipped-no-device`. Use `Status: failed` for any failed proof
-step. Use `Status: blocked` when a safety or access limit stops the attempt before a surface
+`skipped-no-browser-tool`, or `skipped-no-device`. Tell it to register no browser server
+and to report `skipped-no-browser-tool` when no browser tool is loaded in this session. Use
+`Status: failed` for any failed proof step. Use
+`Status: blocked` when a safety or access limit stops the attempt before a surface
 outcome.
 Wait for the proof subagent to finish. For `Status: blocked`, stop onboarding and report the
 blocker. For `Status: failed`, classify the cause before retrying.
@@ -118,31 +120,41 @@ Continue only if the proof result starts with `Status: passed`.
 Run the protected-path check again after the final proof result passes. Apply the
 protected-path rules. Continue only if the audit starts with `Status: passed`.
 
-Do not block core proof on CopilotKit Skills or MCP configuration. Tell the developer that
-the skills install writes a `.agents/skills` directory and `.claude/skills` links into the
-working tree before you run it, because both show up in `git status` and this run cannot
-know whether the project keeps them in version control. Try these tools after the
-application passes proof. Report each tool result separately from the proof result.
+Do not block core proof on the CopilotKit Skills. Tell the developer that the skills
+install writes a `.agents/skills` directory and `.claude/skills` links into the working
+tree before you run it, because both show up in `git status` and this run cannot know
+whether the project keeps them in version control. Install them after the application
+passes proof. Do not register an MCP server: the Skills read the CopilotKit documentation
+without one. Report the Skills result separately from the proof result.
 
 Report the documentation gap and each assumption with the proof evidence. Do not claim
 that the selected documentation proved an inferred step.
 
-When the proof is complete, run `npx --prefer-offline --yes copilotkit@4.18.0 onboard complete`, carrying the
+Record what the proof returned before you complete:
+
+```text
+npx --prefer-offline --yes copilotkit@4.20.0 onboard proof --step round-trip --outcome <passed|failed|skipped>
+```
+
+When the proof is complete, run `npx --prefer-offline --yes copilotkit@4.20.0 onboard complete`, carrying the
 surface-check result the proof subagent returned. Pass exactly one of `--visual-check` or
 `--device-check`, matching this journey's surface:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard complete --visual-check <outcome>
-npx --prefer-offline --yes copilotkit@4.18.0 onboard complete --device-check <outcome>
+npx --prefer-offline --yes copilotkit@4.20.0 onboard complete --visual-check <outcome>
+npx --prefer-offline --yes copilotkit@4.20.0 onboard complete --device-check <outcome>
 ```
 
 `--visual-check` is for a web frontend and takes `performed`, `skipped-no-browser-tool`, or
 `failed`. `--device-check` is for React Native and takes `performed`, `skipped-no-device`,
-or `failed`. Anything but `performed` ends this run as blocked, and the command's output
-names the evidence that is missing. Report it that way.
+or `failed`. `failed` and `skipped-no-device` end this run as blocked, and the command's
+output names the evidence that is missing. Report it that way. `skipped-no-browser-tool`
+completes the run when the round trip was recorded as `passed`, and the output names what
+no browser checked. Without that record it blocks.
 
 If the proof passed and something after it still blocked this run, add
 `--blocked-by <cause>` to the same command, with one of `inspector`, `plan-excluded-capability`,
-or `other`. It ends the run as blocked and names what the blocker leaves unverified. The
-managed Intelligence dashboard is not a cause, because nothing in this graph asks a run to
-open it.
+`browser-control-unresponsive`, or `other`. Pass `browser-control-unresponsive` with
+`--visual-check failed`, for a browser tool that stopped answering. It ends the run as
+blocked and names what the blocker leaves unverified. The managed Intelligence dashboard is
+not a cause, because nothing in this graph asks a run to open it.

package/onboarding/prompts/feature/a2ui/implement.md

@@ -18,7 +18,7 @@ without exposing secrets.
 Before you spawn the implementation subagent, take the authorized list from the CLI:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard audit
+npx --prefer-offline --yes copilotkit@4.20.0 onboard audit
 ```
 
 From its result, copy the paths under `Authorized to modify:`. A passed or failed audit with
@@ -37,7 +37,7 @@ still go on the record. Name the file and why the work needs it, then end your t
 for the developer's answer. If they allow it, record their answer before any edit:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard protect --authorize --unplanned --path <path> --reason "<what the developer said>"
+npx --prefer-offline --yes copilotkit@4.20.0 onboard protect --authorize --unplanned --path <path> --reason "<what the developer said>"
 ```
 
 Continue only when the result starts with `Status: passed`. Run the audit again, take the
@@ -49,7 +49,7 @@ feature stop route below.
 After validation and each repair, run:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard audit
+npx --prefer-offline --yes copilotkit@4.20.0 onboard audit
 ```
 
 Continue only when it starts with `Status: passed`. A path under `Authorized to modify:` is
@@ -60,7 +60,7 @@ with the implementation subagent's `Files changed` section. If that section does
 the path, accept the developer's external change:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard protect --accept-external --path <path>
+npx --prefer-offline --yes copilotkit@4.20.0 onboard protect --accept-external --path <path>
 ```
 
 For an env file where the developer placed a requested credential, use
@@ -73,17 +73,11 @@ Run the audit again after each accepted change. If it still fails, or starts
 with `Status: blocked`, route out and stop:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard read feature/stop
-```
-
-When implementation validation passes, report it:
-
-```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard checkpoint --phase build-validated
+npx --prefer-offline --yes copilotkit@4.20.0 onboard read feature/stop
 ```
 
 If validation cannot pass, or this intent needs a prerequisite the app does not have, use
 the feature stop route above without further changes.
 
 Otherwise run
-`npx --prefer-offline --yes copilotkit@4.18.0 onboard read feature/a2ui/proof`.
+`npx --prefer-offline --yes copilotkit@4.20.0 onboard read feature/a2ui/proof`.

package/onboarding/prompts/feature/a2ui/proof.md

@@ -11,7 +11,7 @@ In the real frontend, send a small request that naturally asks the existing agen
 one compact card or control using A2UI. Require a visible rendered A2UI surface or operation
 in the chat, not only streamed tool arguments, an activity skeleton, or a server log. Record
 the input, visible result, app URL, process IDs, and safe stop commands. If browser control
-is unavailable, report that the browser proof is blocked rather than claiming success.
+is unavailable, report `skipped-no-browser-tool`, and do not claim that the card rendered.
 
 Fix proof failures caused by changed files, then repeat the same proof. Do not replace the
 user's agent response with a hard-coded UI.
@@ -27,7 +27,7 @@ still go on the record. Name the file and why the fix needs it, then end your tu
 for the developer's answer. If they allow it, record their answer before any edit:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard protect --authorize --unplanned --path <path> --reason "<what the developer said>"
+npx --prefer-offline --yes copilotkit@4.20.0 onboard protect --authorize --unplanned --path <path> --reason "<what the developer said>"
 ```
 
 Continue only when the result starts with `Status: passed`. Run the audit again, take the
@@ -36,26 +36,15 @@ same handoff and that list. A subagent that already returned cannot pick up cons
 after it was spawned. If you cannot ask, or the developer declines, take the feature stop
 route below.
 
-Report each attempt at the proof as it ends, counting from one:
+Report the outcome of each attempt as it ends, before any repair. The command counts the
+attempts and the repairs:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard checkpoint --phase journey-attempted --attempt 1
-```
-
-Report each repair cycle the same way, counting from one:
-
-```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard checkpoint --phase repair-attempted --attempt 1
-```
-
-After the final attempt, report the gate exactly once:
-
-```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard proof --step round-trip --outcome <passed|failed|skipped>
+npx --prefer-offline --yes copilotkit@4.20.0 onboard proof --step round-trip --outcome <passed|failed|skipped>
 ```
 
 Use `passed` only for a proved A2UI surface, `failed` for an attempted proof that failed,
-and `skipped` when the proof could not run. Then run `npx --prefer-offline --yes copilotkit@4.18.0 onboard audit`.
+and `skipped` when the proof could not run. Then run `npx --prefer-offline --yes copilotkit@4.20.0 onboard audit`.
 Continue only when it starts with `Status: passed`.
 
 If the audit fails, never repair, reset, or revert a protected path. Compare each named path
@@ -70,9 +59,9 @@ the path and the change the audit reports. Run the audit again after each accept
 If the audit still fails, or starts with `Status: blocked`, route out and stop:
 
 ```text
-npx --prefer-offline --yes copilotkit@4.18.0 onboard read feature/stop
+npx --prefer-offline --yes copilotkit@4.20.0 onboard read feature/stop
 ```
 
 When the audit passes, run
-`npx --prefer-offline --yes copilotkit@4.18.0 onboard read feature/complete` with the actual browser-proof
+`npx --prefer-offline --yes copilotkit@4.20.0 onboard read feature/complete` with the actual browser-proof
 outcome.