@agent-native/core 0.152.1 → 0.153.0

package/corpus/templates/crm/.agents/skills/secrets/SKILL.md

@@ -0,0 +1,339 @@
+---
+name: secrets
+description: >-
+  Declaratively register API keys and service credentials a template needs so
+  they appear in the agent sidebar settings UI and the onboarding checklist.
+  Use for any third-party API key (OpenAI, Stripe, Twilio, etc.) and for
+  surfacing OAuth connections in the unified settings UI.
+scope: dev
+metadata:
+  internal: true
+---
+
+# Secrets Registry
+
+## Non-negotiable rule
+
+Never hardcode credential values. Source, docs, tests, fixtures, prompts, seed
+data, and generated extension/app content may mention credential **names** such
+as `OPENAI_API_KEY`, but must not contain real API keys, tokens, webhook URLs,
+signing secrets, OAuth refresh tokens, or private Builder/customer data.
+
+Secret values are supplied at runtime through deployment configuration, the
+encrypted `app_secrets` vault, `saveCredential` / `resolveCredential`, OAuth, or
+`${keys.NAME}` substitution. Examples must use obvious placeholders such as
+`<OPENAI_API_KEY>` or `${keys.SLACK_WEBHOOK}`, not real-looking copied values.
+
+## When to use
+
+Use this for any external credential your template needs: API keys, service
+tokens, webhook secrets. It gives you:
+
+- A sidebar UI entry for each credential (masked input, rotate, test, delete).
+- Automatic onboarding-checklist items for `required: true` secrets.
+- A stable server-side read API (`readAppSecret`) that decrypts values on
+  demand.
+- Validator hooks for health-checking keys before save and from a Test button.
+
+## When NOT to use
+
+- OAuth flows that need to run the full authorization code exchange — use
+  `@agent-native/core/oauth-tokens` directly to save/refresh tokens. The
+  registry can still surface the OAuth connection in the sidebar by
+  registering a secret with `kind: "oauth"` — that just delegates status
+  lookup to oauth-tokens and renders a Connect button, no `app_secrets` row
+  is written.
+- Purely process-level env vars that are never user-facing (e.g. `NODE_ENV`,
+  deployment flags). Those belong in the onboarding `form` method or the
+  `envKeys` list in `core-routes-plugin`.
+
+## Registering a secret
+
+```ts
+// server/plugins/register-secrets.ts
+import { defineNitroPlugin } from "@agent-native/core/server";
+import { registerRequiredSecret } from "@agent-native/core/secrets";
+
+export default defineNitroPlugin(() => {
+  registerRequiredSecret({
+    key: "OPENAI_API_KEY",
+    label: "OpenAI API Key",
+    description: "Used for Whisper transcription of your recordings.",
+    docsUrl: "https://platform.openai.com/api-keys",
+    scope: "user",
+    kind: "api-key",
+    required: true,
+    validator: async (value) => {
+      const res = await fetch("https://api.openai.com/v1/models", {
+        headers: { Authorization: `Bearer ${value}` },
+      });
+      return res.ok
+        ? { ok: true }
+        : { ok: false, error: `OpenAI rejected the key (HTTP ${res.status})` };
+    },
+  });
+});
+```
+
+### OAuth in the unified UI
+
+```ts
+registerRequiredSecret({
+  key: "GOOGLE_CONNECTED",
+  label: "Google account",
+  description: "Grants access to Gmail / Calendar APIs.",
+  scope: "user",
+  kind: "oauth",
+  required: true,
+  oauthProvider: "google", // must match the provider id in oauth-tokens
+  oauthConnectUrl: "/_agent-native/google/auth-url",
+});
+```
+
+The sidebar shows a Connect button instead of a text input; no `app_secrets`
+row is written — status is derived from `hasOAuthTokens("google")`.
+
+## Registered options
+
+| Field              | Type                                    | Purpose                                                                  |
+| ------------------ | --------------------------------------- | ------------------------------------------------------------------------ |
+| `key`              | `string`                                | Env-var style name (`OPENAI_API_KEY`). Also the storage key.             |
+| `label`            | `string`                                | Human-readable title in the sidebar.                                     |
+| `description`      | `string?`                               | Subtitle under the label.                                                |
+| `docsUrl`          | `string?`                               | "Get key" link rendered on the card.                                     |
+| `scope`            | `"user" \| "workspace"`                 | Per-user or shared across the active org.                                |
+| `kind`             | `"api-key" \| "oauth"`                  | Drives UI and storage behavior.                                          |
+| `required`         | `boolean?`                              | When true, an onboarding step is auto-injected.                          |
+| `validator`        | `(v) => Promise<boolean \| {ok,error}>` | Runs on save and from the Test button. Never log `v`.                    |
+| `oauthProvider`    | `string?` (oauth-kind only)             | Provider id in `oauth-tokens` that backs this entry.                     |
+| `oauthConnectUrl`  | `string?` (oauth-kind only)             | URL the Connect button points at.                                        |
+
+## Reading a secret from an action
+
+```ts
+import { z } from "zod";
+import { defineAction } from "@agent-native/core/action";
+import { readAppSecret } from "@agent-native/core/secrets";
+import { getRequestUserEmail } from "@agent-native/core/server";
+
+export default defineAction({
+  description: "Transcribe an audio file with Whisper",
+  schema: z.object({ fileUrl: z.string() }),
+  run: async ({ fileUrl }) => {
+    const email = await getRequestUserEmail();
+    if (!email) throw new Error("Not signed in");
+
+    const stored = await readAppSecret({
+      key: "OPENAI_API_KEY",
+      scope: "user",
+      scopeId: email,
+    });
+    const apiKey = stored?.value;
+    if (!apiKey) {
+      throw new Error(
+        "OPENAI_API_KEY is not set. Configure it in the sidebar settings.",
+      );
+    }
+
+    // …call OpenAI. NEVER log the key or include it in error messages.
+  },
+});
+```
+
+Rules:
+
+- **Never log the value.** The read layer enforces this server-side; your
+  code must do the same.
+- **Use env vars only for deploy-level secrets.** If a credential is
+  user-scoped, org-scoped, or workspace-scoped, read the scoped vault/credential
+  store. Do not add a `process.env` fallback that makes every user inherit one
+  deployment's key.
+- **Scope matches the registration.** `scope: "user"` → pass the user email.
+  `scope: "workspace"` → pass the active `orgId` from
+  `getOrgContext(event).orgId`.
+- **One resolver per key, and every runtime path goes through it.** Before
+  reading a credential, grep the app for the key name. If a resolver already
+  exists (`resolveXConfig`, a connector, a client factory), call it — do not
+  read `process.env` for that key a second time somewhere else.
+- **Identity comes from the caller, not the module.** A shared helper under
+  `server/lib/` takes the email as a parameter; only an entrypoint (action,
+  route, cron) decides whose identity it is. A library that resolves its own
+  identity from `process.env.AGENT_USER_EMAIL` authorizes the deployment rather
+  than the caller.
+
+The failure this prevents is not a leak — it is a split brain. When the config
+check reads the vault and the feature reads `process.env`, the settings UI and
+onboarding checklist report the integration as **configured** while every actual
+call fails with "env var is required". The credential is right there in the
+vault, so the error names the wrong cause and sends everyone hunting for a sync
+or redeploy problem that does not exist. This has now shipped four times in one
+app (BigQuery, Jira, Pylon, Academy).
+
+Before finishing any change that touches a credential, run the guard from the
+app directory — it finds this whether the app lives in `templates/` or in a
+workspace repo:
+
+```bash
+npx agent-native doctor --only no-env-credentials
+```
+
+## `resolveCredential` sees exactly one organization
+
+`resolveCredential(key, { userEmail, orgId })` checks the user scope, then the
+one `orgId` you pass, then the solo workspace. That is the whole search. It is
+correct for a signed-in request, and wrong for two common cases:
+
+- **The caller has no organization.** A cron, a scheduled job, or any CLI run
+  without a real member identity resolves no `orgId`, so only the user and solo
+  scopes are ever consulted — and a shared key is in neither.
+- **The key was synced under a different organization.** `app_secrets` has no
+  scope visible to every org (`readAppSecret` is strict equality on
+  `(scope, scope_id, key)`), so a key the vault UI advertises as available to
+  every app is unreadable from any org except the one that ran the sync.
+
+Both produce the same misleading "not set" that the split brain above produces,
+which is why swapping `process.env` for `resolveCredential` can look like a fix
+and change nothing. A workspace app should read shared keys through a resolver
+that also sweeps the caller's other memberships and a designated vault org —
+see `resolveConnectorSecret` and `AGENT_VAULT_ORG_ID` in the builder-workspace
+repo for the shape, including the boot-time assertion that the deployment really
+is single-tenant before a non-membership-gated fallback is safe.
+
+**The doctor guard does not catch this second form** — it looks for
+`process.env` reads, and `resolveCredential` is not one. Grep for
+`resolveCredential` yourself and confirm each call sits somewhere a single-org
+lookup is genuinely the right question. Findings that pair it with a
+`?? process.env.KEY` fallback are the sanctioned deploy-level escape hatch; the
+bugs are calls whose value can only live in another org's vault.
+
+## HTTP routes
+
+Core routes plugin mounts these under `/_agent-native/secrets/` automatically:
+
+- `GET /_agent-native/secrets` — list registered secrets with status (`set`
+  / `unset` / `invalid`), metadata, and — for set api-keys — the last 4
+  characters. Values are never returned.
+- `POST /_agent-native/secrets/:key` — body `{ value, scope?, scopeId? }`.
+  Runs the registered validator; returns 400 with the error on failure.
+- `DELETE /_agent-native/secrets/:key` — remove the stored value.
+- `POST /_agent-native/secrets/:key/test` — re-run the validator against the
+  currently stored value.
+
+## Storage & encryption
+
+- Values are stored in `app_secrets` (created on-demand; no migration
+  needed).
+- Values are encrypted at rest with AES-256-GCM. Generic app-local secrets
+  prefer `<APP_NAME>_SECRETS_ENCRYPTION_KEY` (for example,
+  `ANALYTICS_SECRETS_ENCRYPTION_KEY`), then `SECRETS_ENCRYPTION_KEY`, then
+  `BETTER_AUTH_SECRET`.
+- Workspace-shared `app_secrets` prefer
+  `WORKSPACE_SECRETS_ENCRYPTION_KEY`, then the legacy shared
+  `SECRETS_ENCRYPTION_KEY`, then a purpose-derived key from `A2A_SECRET`.
+  Better Auth and app-scoped keys remain legacy read candidates; a successful
+  legacy decrypt is compare-and-swap migrated to the preferred shared key
+  without changing the row timestamp.
+- Set the same stable `WORKSPACE_SECRETS_ENCRYPTION_KEY` in every app that
+  reads a shared vault. It is intentionally separate from app-local OAuth
+  encryption. If the vault still relies on `A2A_SECRET`, rotating A2A material
+  also rotates its encryption key. Add the stable vault key everywhere and
+  read/migrate existing rows before rotating A2A.
+- To rotate the dedicated workspace key itself, deploy the new value together
+  with `WORKSPACE_SECRETS_ENCRYPTION_KEY_PREVIOUS=<old value>`, let reads
+  migrate ciphertext, then remove the previous key after the migration window.
+- If no configured key material exists, development uses a machine-local
+  fallback and logs a one-time warning. Production fails closed.
+
+## Ad-hoc Keys
+
+Ad-hoc keys are user-created secrets that are not declared by the template.
+Users create them through the settings UI or the agent chat to use with
+automations and the `web-request` tool. They support `${keys.NAME}`
+substitution in outbound HTTP requests.
+
+### Ad-hoc API
+
+Core routes plugin mounts these under `/_agent-native/secrets/adhoc`:
+
+- `GET /_agent-native/secrets/adhoc` — list all ad-hoc keys (name, last 4
+  chars, URL allowlist). Values are never returned.
+- `POST /_agent-native/secrets/adhoc` — body `{ name, value, urlAllowlist? }`.
+  Creates or updates an ad-hoc key.
+- `DELETE /_agent-native/secrets/adhoc/:name` — remove an ad-hoc key.
+
+### URL Allowlists
+
+Each ad-hoc key can have a URL allowlist — an array of origin URLs that
+restrict where the key's value can be sent. The check is origin-level
+(scheme + host + port). If no allowlist is configured, the key can be used
+with any URL.
+
+```ts
+// Creating a key with an allowlist
+POST /_agent-native/secrets/adhoc
+{
+  "name": "SLACK_WEBHOOK",
+  "value": "<SLACK_WEBHOOK_URL_FROM_SETTINGS>",
+  "urlAllowlist": ["https://hooks.slack.com"]
+}
+```
+
+### `${keys.NAME}` Substitution
+
+The `web-request` tool supports `${keys.NAME}` placeholders in the URL,
+headers, and body. Substitution happens server-side after the agent emits
+the tool call — the raw secret value never enters the agent's context.
+
+```ts
+import {
+  resolveKeyReferences,
+  validateUrlAllowlist,
+} from "@agent-native/core/secrets/substitution";
+
+// Resolve all ${keys.NAME} references in a string
+const { resolved, usedKeys } = await resolveKeyReferences(
+  "Bearer ${keys.API_TOKEN}",
+  "user",
+  "user@example.com",
+);
+
+// Validate a URL against a key's allowlist
+const allowed = validateUrlAllowlist(
+  "https://hooks.slack.com/services/<WORKSPACE>/<CHANNEL>/<SECRET>",
+  ["https://hooks.slack.com"],
+);
+```
+
+Key resolution falls back from user scope to workspace scope, so users can
+override shared keys without breaking automations that reference workspace
+defaults.
+
+## Dispatch Vault Access
+
+Dispatch workspaces have a vault access policy for workspace app credentials:
+
+- `all-apps` is the default. Every saved Dispatch vault key is available to
+  every workspace app; `sync-vault-to-app` pushes all vault keys to the target
+  app.
+- `manual` requires explicit per-app grants. Use
+  `create-vault-grant` / `grant-vault-secrets-to-app`, then
+  `sync-vault-to-app`.
+
+Use `get-vault-access-settings` before deciding whether to create grants, and
+use `set-vault-access-settings` only when the user asks to change the policy.
+
+### Key Files (ad-hoc)
+
+| File                                           | Purpose                                     |
+| ---------------------------------------------- | ------------------------------------------- |
+| `packages/core/src/secrets/substitution.ts`    | `resolveKeyReferences()`, `validateUrlAllowlist()` |
+| `packages/core/src/tools/fetch-tool.ts`        | `web-request` tool consuming key references |
+
+## Related skills
+
+- `onboarding` — the setup checklist that required secrets show up in.
+- `actions` — where you'll read secrets when calling third-party APIs.
+- `authentication` — session scoping; `scope: "user"` uses the session
+  email.
+- `security` — input validation and never logging secrets.
+- `automations` — ad-hoc keys power `${keys.NAME}` in automation web requests.

package/corpus/templates/crm/.agents/skills/security/SKILL.md

@@ -0,0 +1,324 @@
+---
+name: security
+description: >-
+  Secure coding practices for agent-native apps: input validation, SQL
+  injection, XSS, secrets, data scoping, and auth. Use when writing any action,
+  route, or component that touches user data or external input.
+scope: dev
+metadata:
+  internal: true
+---
+
+# Security
+
+## Rule
+
+Use the framework's security primitives everywhere. Never bypass them.
+
+## Absolute Secrets Rule
+
+Never hardcode secret values or real private data. This applies to source code,
+docs, tests, fixtures, generated prompts, screenshots, seed data, and extension
+HTML just as much as production code.
+
+Do not paste or invent real-looking API keys, bearer tokens, OAuth refresh
+tokens, webhook URLs, signing secrets, private Builder/internal data, or customer
+data into the repo. Examples must use obvious placeholders such as
+`<OPENAI_API_KEY>`, `${keys.SLACK_WEBHOOK}`, `sk-test-example`, or
+`example.customer@example.com`. Test literals should be clearly fake and must
+not match real provider token formats when an `example` token will do.
+
+Credential values enter the system only through approved runtime channels:
+deployment env vars for deploy-level secrets, the encrypted `app_secrets` vault
+or `saveCredential` / `resolveCredential` for user/org/workspace API keys, and
+`oauth_tokens` for OAuth. Code and instructions may name the credential key
+(`OPENAI_API_KEY`), but must never contain the credential value.
+
+## Input Validation
+
+Use `defineAction` with a Zod `schema:` for every action. The framework validates input automatically and returns clear 400 errors for HTTP callers and structured error results for agent tool calls.
+
+```ts
+export default defineAction({
+  schema: z.object({
+    email: z.string().email(),
+    role: z.enum(["admin", "member"]),
+    limit: z.coerce.number().int().min(1).max(100).default(25),
+  }),
+  run: async (args) => { /* args is fully typed and validated */ },
+});
+```
+
+The legacy `parameters:` field (plain JSON Schema) has no runtime validation — do not use it for new code.
+
+## Large Payloads
+
+Do not accept or persist unbounded base64/file blobs through actions, SQL writes,
+or `application_state`. Route uploads through the file-upload provider and store
+references. This prevents database bloat, slow hot paths, and accidental secret
+or customer-data embedding inside binary payloads.
+
+## SQL Injection
+
+Never concatenate user input into SQL strings. Use Drizzle ORM's query builder (always safe) or parameterized queries:
+
+```ts
+// Safe — Drizzle ORM
+await db.select().from(users).where(eq(users.email, args.email));
+
+// Safe — parameterized raw SQL
+await client.execute({ sql: "SELECT * FROM users WHERE id = ?", args: [id] });
+
+// NEVER do this
+await client.execute(`SELECT * FROM users WHERE id = '${id}'`);
+```
+
+## XSS
+
+- React auto-escapes JSX content — trust it.
+- Never use `dangerouslySetInnerHTML`, `innerHTML`, `eval()`, or `document.write()` with user-controlled content.
+- For rich text editing, use TipTap (framework dependency).
+- For rendering markdown, use `react-markdown`.
+
+## SSRF
+
+Any server-side `fetch` of a user- or agent-controlled URL must go through the framework SSRF guard — a bare `fetch()` can be steered at cloud metadata (`169.254.169.254`), `localhost`, or internal services.
+
+```ts
+import { ssrfSafeFetch } from "@agent-native/core/extensions/url-safety";
+// Blocks private/internal targets, re-checks the resolved IP at connect time
+// (DNS rebinding), and re-validates every redirect hop.
+const res = await ssrfSafeFetch(userProvidedUrl, {}, { maxRedirects: 3 });
+```
+
+For a pre-flight-only check (e.g. before a streaming or one-shot fetch), use `isBlockedExtensionUrlWithDns(url)` plus `createSsrfSafeDispatcher()` from the same module, and set `redirect: "manual"`. Never let the default `fetch` follow redirects for an untrusted URL — a public URL can 30x into the private network.
+
+## Secrets
+
+- OAuth tokens go in the `oauth_tokens` store via `saveOAuthTokens()`.
+- Per-user / per-org API keys go through `saveCredential` / `resolveCredential` (`@agent-native/core/credentials`) or the `app_secrets` vault. Both encrypt values at rest with AES-256-GCM (keyed by `SECRETS_ENCRYPTION_KEY`, falling back to `BETTER_AUTH_SECRET`; production refuses to start without one).
+- Never hand-roll secrets into `settings`, `application_state`, source code, or action responses sent to the client. The credential / vault APIs above are the only sanctioned stores.
+- Never commit real keys, tokens, webhook URLs, signing secrets, or private
+  Builder/customer data in examples or fixtures. Use placeholders that cannot be
+  mistaken for working credentials.
+
+## User Credentials Are Per-User Data — Never `process.env`
+
+User credentials (API keys, third-party tokens) are per-user (or per-org) data. They MUST live in SQL, scoped per-user (`u:<email>:credential:KEY`) or per-org (`o:<orgId>:credential:KEY`). Always read with the request context:
+
+```ts
+import { resolveCredential } from "@agent-native/core/credentials";
+const apiKey = await resolveCredential("OPENAI_API_KEY", { userEmail, orgId });
+```
+
+Values are encrypted at rest (AES-256-GCM, shared `secrets/crypto.ts`): `saveCredential` encrypts on write and `resolveCredential` decrypts on read, with a transparent fallback for legacy plaintext rows. The agent's raw `db-query` / `db-exec` tools also cannot read credential rows — they are excluded from the scoped `settings` view. To encrypt pre-existing rows in place, run `pnpm action db-migrate-encrypt-credentials` (idempotent, non-destructive; needs the same `SECRETS_ENCRYPTION_KEY` / `BETTER_AUTH_SECRET` as the app).
+
+On 2026-04-29 the previous one-arg `resolveCredential(key)` form fell back to `process.env[key]` and an unscoped global `settings` row, so every signed-in user inherited the deployment's credentials. Two guards now block this in CI (`pnpm prep`):
+
+- `scripts/guard-no-env-credentials.mjs` — bans `process.env.<KEY>` reads in `packages/core/src/credentials/`, `secrets/`, `vault/`, and `templates/*/server/{lib,routes/api}/credential*` paths, except for an explicit allowlist of deploy-level vars (`DATABASE_URL`, `BETTER_AUTH_SECRET`, `NETLIFY_*`, etc.). Per-line opt-out: `// guard:allow-env-credential — <reason>`.
+- `scripts/guard-no-unscoped-credentials.mjs` — bans one-arg calls to `resolveCredential` / `hasCredential` / `saveCredential` / `deleteCredential`. Per-line opt-out: `// guard:allow-unscoped-credential — <reason>`.
+
+If a deploy-level value genuinely needs an env var (CI-set token, host secret), it's not a user credential — keep it out of the credentials/ secrets/ vault/ paths and the env-credentials guard won't see it.
+
+## Guards
+
+Two more CI guards (also wired into `pnpm prep`) target the 2026-04 cross-tenant leak class — request-state escaping into shared process state, and dev-mode sentinel identities used as production fallbacks.
+
+- `scripts/guard-no-env-mutation.mjs` — bans `process.env.<KEY> = …` (and bracket / compound forms) anywhere in production code. On serverless, every warm container handles many concurrent requests in one Node process, so `process.env` mutation leaks across in-flight requests (the "restore" line at the end of a handler races and never helps — most recently the Zoom webhook). Use `runWithRequestContext({ userEmail, orgId, timezone }, fn)` from `@agent-native/core/server` instead — it's AsyncLocalStorage-backed and per-request safe. Allowlisted paths: `scripts/`, `*.spec.ts` / `*.test.ts`, `packages/core/src/dev**`, `templates/*/test/`, anything under `/cli/` or `/scaffold/`. Per-line opt-out: `process.env.X = y // guard:allow-env-mutation — <reason>`.
+- `scripts/guard-no-localhost-fallback.mjs` — bans the literal `"local@localhost"` / `'local@localhost'` / `` `local@localhost` `` in production code. The bug class: `getRequestUserEmail() ?? "local@localhost"` silently pools every unauthenticated request into a single shared tenant, leaking credentials, tools, and `application_state` rows between accounts. The right behavior is to throw / 401 when there's no session. Allowlisted paths: the dev-mode auth shim (`packages/core/src/server/auth.ts`), `packages/core/src/dev**`, tests, `scripts/`, `seed/` / `seeds/`, plus a few framework helpers that intentionally inspect or migrate the dev identity. SQL DDL `DEFAULT 'local@localhost'` and the Drizzle helper `.default('local@localhost')` are skipped per-line — schema column defaults are intentional dev fixtures, not the dangerous fallback pattern. Per-line opt-out: `email ?? "local@localhost" // guard:allow-localhost-fallback — <reason>`.
+
+  The same guard also bans the **ambient process identity** as a request-scoped fallback: `?? process.env.AGENT_USER_EMAIL`, `|| process.env.WORKSPACE_OWNER_EMAIL`, and `?? getAmbientUserEmail()`. Those name the identity of the *deployment*, not the caller, so a handler reading one authorizes whoever the deploy env names rather than whoever signed in — it fails open toward more privilege. Exempt paths are the ones with no request by construction: `scripts/` (including `templates/*/scripts/`), `src/cli/`, `src/scripts/`, `script-helpers.ts` / `script-entries.ts`, tests, and seeds. It deliberately matches only the `??` / `||` fallback position, so reading the same env var to build an admin allowlist (`envEmails("WORKSPACE_OWNER_EMAIL").includes(email)`) is untouched. Known gaps, both requiring dataflow analysis the guard does not do: an env read aliased through a local (`const fallback = process.env.AGENT_USER_EMAIL; … ?? fallback`) and a destructured `const { AGENT_USER_EMAIL } = process.env`.
+
+## Auth
+
+- All actions are protected by the auth guard automatically.
+- Prefer actions for normal app data. Do not hand-write `/api/*` routes for
+  CRUD, data queries, or action re-exports just to add auth; action endpoints
+  already get auth and request context.
+- If you must create custom `/api/` routes, always call `getSession(event)` and reject requests without a session:
+
+```ts
+import { getSession } from "@agent-native/core/server";
+
+export default defineEventHandler(async (event) => {
+  const session = await getSession(event);
+  if (!session) throw createError({ statusCode: 401 });
+  // ...
+});
+```
+
+- Never create unprotected routes that modify data.
+
+**Exception — the SSR HTML/`.data` catch-all is deliberately session-blind.**
+The rule above is for routes that read or mutate user data. The SSR page
+render and React Router `.data` route are different: they serve one
+impersonal, public-cacheable shell to every visitor by design, and loaders on
+that path render no user data. Do not "fix" this by adding `getSession`,
+`private`, or `no-store` to it — that regresses the CDN cache contract for the
+whole site. Data scoping lives in actions and API routes; the client gates
+private UI after the shell loads. See the `authentication` skill and
+`guard:ssr-cache-shell` / `ssr-handler.spec.ts`.
+
+**Authorization beyond "is there a session" — `authorize`.** The auth guard only
+proves *someone* is signed in. To restrict an operation to some teammates, set
+`authorize` on the `defineAction`. It wraps `run`, so it applies at every
+dispatch site (agent tool, HTTP, frontend, MCP, A2A, CLI) — unlike
+`needsApproval`, which is honored by the agent loop and MCP 2026 hosts that
+support elicitation, but is not an authorization boundary for every dispatch
+site.
+
+```ts
+import { coachAccess } from "../lib/access.js"; // defineAppRoles(...)
+
+export default defineAction({
+  description: "Archive a client roster.",
+  schema: z.object({ id: z.string() }),
+  authorize: coachAccess.requireAny("coach-admin"),
+  run: async (args) => {
+    /* ... */
+  },
+});
+```
+
+A guard that throws denies with its own message; returning `false` denies
+generically; anything else (including `undefined`) allows. Guarded actions need
+a user identity — an unattended CLI/cron caller with no user email is denied.
+
+`authorize` is **not** a substitute for `accessFilter` / `assertAccess`. It
+decides whether this caller may perform the operation at all; `accessFilter` /
+`assertAccess` scope which rows a permitted caller may see or touch. A
+restricted write action needs both. See the `authentication` skill for
+`defineAppRoles` and the `actions` skill for the full surface.
+
+## Human-in-the-Loop Approval for High-Consequence Actions
+
+For a small set of outward-facing, hard-to-undo operations — sending an email, charging a card, deleting an account, posting publicly — auth and access control are necessary but not sufficient: you also do not want the **agent** to perform them autonomously. Set `needsApproval` on the `defineAction` so the agent cannot run the action without a human approving the specific call.
+
+```ts
+export default defineAction({
+  description: "Send an email via Gmail.",
+  schema: z.object({ to: z.string(), subject: z.string(), body: z.string() }),
+  needsApproval: true, // or (args, ctx) => boolean | Promise<boolean>
+  run: async (args) => {
+    /* ...actually send... */
+  },
+});
+```
+
+When the gate is truthy and the call is not yet approved, the agent loop emits
+an `approval_required` event and **stops the turn — `run()` never executes**.
+The human approves via the chat UI's Approve affordance, which re-issues the
+turn with the call's stable `approvalKey`; only then does the action run. MCP
+2026 hosts receive an `input_required` approval request bound to the exact
+caller, action, and arguments. The signed response state is backed by a durable,
+single-use grant, so denial, expiry, replay, missing host support, or invalid
+state fails closed before `run()`. A predicate gates conditionally (e.g. only
+external recipients) and **fails closed** — a throw is treated as "approval
+required".
+
+Rules:
+
+- Reach for `needsApproval` only for genuinely high-consequence operations. The default is off, and the framework intentionally keeps approvals rare — over-gating turns the agent into a click-through wizard. The canonical (and intentionally lone) framework example is Mail's `send-email`.
+- `needsApproval` is **not** a substitute for `accessFilter` / `assertAccess` or for hiding sensitive operations from the model with `agentTool: false` / `toolCallable: false`. It is the layer for "a human must explicitly bless this specific outward-facing call," not for scoping data. See the `actions` skill for the full surface.
+
+## Custom HTTP Routes Must Apply Access Control Themselves
+
+This is the single most-failed rule in the codebase. Auto-mounted action routes (`/_agent-native/actions/...`) get a request context wired up automatically. **Hand-written `/api/*` Nitro routes do not.** If your handler queries an ownable resource (any table with `...ownableColumns()`), you MUST:
+
+1. Read the session: `const session = await getSession(event).catch(() => null)`.
+2. Run the work inside `runWithRequestContext({ userEmail: session?.email, orgId: session?.orgId }, fn)` from `@agent-native/core/server`.
+3. Inside `fn`, query through one of:
+   - `accessFilter(table, sharesTable)` in the WHERE clause for list/read-many.
+   - `resolveAccess("<type>", id)` for read-by-id (returns null if no access — return 404, not 403, so existence isn't leaked).
+   - `assertAccess("<type>", id, "viewer"|"editor"|"admin")` for write/delete-by-id.
+
+```ts
+// Bad — Brent's signup leaked every other user's decks because of this exact shape.
+export default defineEventHandler(async () => {
+  const db = getDb();
+  return db.select().from(schema.decks); // no access filter!
+});
+
+// Good
+import { getSession, runWithRequestContext } from "@agent-native/core/server";
+import { accessFilter } from "@agent-native/core/sharing";
+export default defineEventHandler(async (event) => {
+  const session = await getSession(event).catch(() => null);
+  return runWithRequestContext(
+    { userEmail: session?.email, orgId: session?.orgId },
+    async () => {
+      const db = getDb();
+      return db
+        .select()
+        .from(schema.decks)
+        .where(accessFilter(schema.decks, schema.deckShares));
+    },
+  );
+});
+```
+
+`scripts/guard-no-unscoped-queries.mjs` runs in `pnpm prep` and fails the build if any file in `templates/*/server/`, `templates/*/actions/`, or `packages/*/src/` queries an ownable table without one of the access helpers. Last-resort opt-out is the marker comment `// guard:allow-unscoped — <reason>` — only use it for cases like the sharing primitives themselves or share-token-public viewer endpoints, and always include a reviewer-readable reason.
+
+## Data Scoping
+
+In production, the framework automatically restricts all agent SQL queries to the current user's data using temporary views. This is enforced at the SQL level — the agent cannot bypass it.
+
+The `db-query` / `db-exec` tools (and the extension SQL bridge, which shares the same path) reject schema-qualified table references like `public.<table>` or `main.<table>` — a qualified name resolves to the base table and would skip the temp view. Use bare table names; scoping is applied automatically.
+
+### Per-User Scoping (`owner_email`)
+
+Every template table with user data **must** have an `owner_email` text column:
+
+1. Framework detects `owner_email` via schema introspection
+2. Creates temp views `WHERE owner_email = <current user>` before each query
+3. Auto-injects `owner_email` into INSERT statements
+
+The current user is resolved from `AGENT_USER_EMAIL` (set automatically from the session).
+
+### Per-Org Scoping (`org_id`)
+
+For multi-org apps, tables also need `org_id`:
+
+1. `WHERE org_id = <current org>` is added (in addition to `owner_email` if present)
+2. `org_id` is auto-injected into INSERT statements
+
+Enable org scoping in the agent-chat plugin:
+
+```ts
+createAgentChatPlugin({
+  resolveOrgId: async (event) => {
+    const ctx = await getOrgContext(event);
+    return ctx.orgId;
+  },
+});
+```
+
+### Column Conventions
+
+| Column        | Purpose                 | Required                        |
+| ------------- | ----------------------- | ------------------------------- |
+| `owner_email` | Per-user data isolation | Yes, for all user-facing tables |
+| `org_id`      | Per-org data isolation  | Yes, for multi-org apps         |
+
+Run `pnpm action db-check-scoping` to verify. Use `--require-org` for multi-org apps.
+
+## Checklist
+
+- [ ] New action uses `defineAction` with a Zod `schema:`
+- [ ] No SQL string concatenation with user input
+- [ ] No `dangerouslySetInnerHTML` with user content
+- [ ] Server-side fetches of user/agent URLs use `ssrfSafeFetch`, not bare `fetch`
+- [ ] Secrets stored via `saveCredential` / the vault (encrypted), never raw in `settings` or responses
+- [ ] No hardcoded API keys, tokens, webhook URLs, signing secrets, real
+      credential-looking strings, private Builder/internal data, or customer data
+- [ ] New env vars in `.env` only, not committed
+- [ ] New user-data tables have `owner_email` column
+- [ ] Custom routes call `getSession` and reject unauthenticated requests
+- [ ] Actions only some teammates may run set `authorize` (in addition to any
+      `accessFilter` / `assertAccess` row scoping)
+
+## Related Skills
+
+- `storing-data` — SQL patterns and the agent's db tools
+- `actions` — `defineAction` with Zod schema validation
+- `authentication` — Auth modes, sessions, and org context