@nylorun/runtime 0.19.1-beta → 0.20.0-beta

package/CHANGELOG.md

@@ -1,5 +1,31 @@
 # Changelog
 
+## 0.20.0-beta
+
+### Major Changes
+
+- cf5eb9c: **Protocol 9: the Runtime API is an OAuth 2.1 resource server for your identity provider's tokens.** Nylorun is still never the authorization server. Upgrade every package together; MIGRATION.md has the details.
+
+  - **Breaking (`@nylorun/runtime`): a refused credential is `401`, not the opaque `404`.** No `Authorization` is `401 credential_required`; an unknown key, or a token no trusted issuer signed or that fails a check, is `401 credential_invalid`. Each carries `WWW-Authenticate: Bearer` (OAuth 2.1 §5.3), naming the protected resource metadata (`resource_metadata`) on the Runtime API when the Runtime has trusted issuers; the Management API's challenge is a bare `Bearer`. `token_expired` and `issuer_unavailable` gain the challenge too. Another Tenant named, a Tenant that could not open and an unsigned capability link stay the opaque `404`.
+  - **Breaking (`@nylorun/runtime`): a request with neither `Nylorun-Protocol` nor `Authorization`** on an API route gets that route's `401` challenge instead of `426`, so a generic OAuth client learns where to sign in.
+  - `@nylorun/runtime`: a token without a route's scope still gets `403 scope_required`, now with `WWW-Authenticate: Bearer error="insufficient_scope", scope="…"`. The `Bearer` scheme is matched in any case (RFC 9110).
+  - `@nylorun/runtime`: `GET /.well-known/oauth-protected-resource` (RFC 9728) serves the resource (`NYLORUN_PUBLIC_URL`, else the request's origin), the identity file's issuers in order and their scopes, with no key or protocol; `404` without an identity file.
+  - **Breaking (`@nylorun/runtime`): the identity file drops `maxLifetime`.** The issuer sets its tokens' lifetimes; `exp` is required, `iat` no longer is. A key the file does not define, `maxLifetime` included, is ignored and logged (`identity_file_key_ignored`) instead of stopping the boot. `subject` defaults to `{sub}`, `scopes` to `{ claim: scope }` and `allowedScopes` to every token scope but `studio`.
+  - `@nylorun/core`: `PROTOCOL_VERSION` is 9 and `HOST_PROTOCOL` 4–9, with the required feature `resource-server`; `ERROR_CODES` adds `credential_required` and `credential_invalid`. `@nylorun/agents`, `@nylorun/admin` and `@nylorun/cli` send protocol 9.
+
+### Patch Changes
+
+- 6711320: Cancels and sessions resets reach the other Runtime processes over a control bus on Postgres, not S2's `tenant/control` stream, which is gone: nothing inside the Runtime depends on S2 any more. A cancel writes a `session.cancel` signal in its own transaction and Postgres notifies every process at commit (`LISTEN`/`NOTIFY` on `nylorun_control`), so a cancel aborts the Worker's model or tool call even while S2 is down. Each process holds one `LISTEN` connection of its own and reads recent signals back every 5 s and after a reconnect; connect the Runtime to Postgres directly or through a pooler in session mode. Migration `0015_control_signals` adds `nylorun.control_signals`; the Tenant sweep deletes signals older than an hour. While processes of the previous beta still run beside new ones, a cancel between them is stopped by the Session Store check before the next effect, as when a signal was lost.
+- 6711320: `/ready` no longer checks S2: its `checks` cover the listener, the Tenant, Postgres and Restate, and an unreachable S2 leaves it `200`. S2 only serves API listeners (history, SSE, AG-UI and A2A), so an outage degrades those reads and never makes the Runtime unready. S2's reachability stays in the Tenant's status (`GET /v1/tenant`, `streams.reachable`).
+- 2ed5fe0: The event source `sandboxd` is now `sandboxes`, the name the sandbox lifecycle service took on 3 Oct, matching core, loop, gates, keys, egress and harness. Nothing emits it yet, so no stored event changes; the OpenAPI document's `EventSourceKind` enum shows the new name, and `--service sandboxes` is the blueprint service this release refuses as not yet shipped.
+- Pin core to the tested release.
+- Pin harness to the tested release.
+- Updated dependencies [cf5eb9c]
+- Updated dependencies [2ed5fe0]
+- Updated dependencies
+  - @nylorun/core@0.16.0-beta
+  - @nylorun/harness@0.23.1-beta
+
 ## 0.19.1-beta
 
 ### Patch Changes

package/README.md

@@ -136,7 +136,7 @@ The Host root is `NYLORUN_HOME` or `~/.nylorun` (for a local Tenant,
 | Route | Auth | Notes |
 | --- | --- | --- |
 | `GET /health` | none | `service: "nylorun-runtime"`, `hostId`, protocol `{min,max,features}`, pid |
-| `GET /ready` | none | Listener up, the Tenant open, and Postgres, Restate and S2 answer (`checks`); `harness: { mode, connected }` while the Tenant is open |
+| `GET /ready` | none | Listener up, the Tenant open, and Postgres and Restate answer (`checks`). S2 is not checked: its health is in `GET /v1/tenant` (`streams.reachable`); `harness: { mode, connected }` while the Tenant is open |
 | `GET /openapi/runtime.json` | none | The Runtime API's OpenAPI 3.2 document (below); alias `GET /openapi.json`. Refuses an `Origin` |
 | `GET /openapi/management.json` | none | The Management API's OpenAPI 3.2 document. Refuses an `Origin` |
 | `/v1/*` Runtime API routes (all but `/v1/tenant/*`) | application key or trusted issuer's token | Require `Nylorun-Protocol`; nothing names the Tenant. A management key is `403 key_role_mismatch`, except on `/v1/me` and the public `GET /v1/access/jwks` |
@@ -154,15 +154,19 @@ presents a trusted issuer's token; keys (application and management)
 are refused from browsers before they are looked up. The Runtime sends no CORS headers, and answers
 `OPTIONS` with `204` and `Allow` only: the operator's reverse proxy answers preflights
 ([DEPLOYMENT.md](../DEPLOYMENT.md#calling-the-runtime-from-browsers-and-apps)). Missing or
-unsupported protocol → `426` before authentication (the Host serves protocols 4 to 8;
-protocol 8 gives keys roles, moves vaults and signing keys under `/v1/tenant` and removes
+unsupported protocol → `426` before authentication, unless the request sends neither
+`Nylorun-Protocol` nor `Authorization` (the Host serves protocols 4 to 9; protocol 9 makes the
+Runtime API an OAuth 2.1 resource server: a missing or rejected credential is `401
+credential_required` or `credential_invalid` with a `WWW-Authenticate: Bearer` challenge, a
+token without a route's scope gets `error="insufficient_scope"`, and
+`/.well-known/oauth-protected-resource` lists the trusted issuers (RFC 9728); protocol 8 gives keys roles, moves vaults and signing keys under `/v1/tenant` and removes
 the Admin API, and those old paths answer `404` with no alias; protocol 7 removes subject tokens, the access policy, revocations, browser keys and
 derived principals, whose routes answer `404`; protocol 6 adds file and folder artifacts and message `parts`; at each turn's end the
 Runtime exports `/workspace/outputs` of the session's sandbox as a version of its
 `outputs` folder, read at `/v1/artifacts/{id}/versions/{n|latest}/tree`, `/files/{path}`,
 `/diff?from=` and `/zip`). Protocol 5 and later clients name no Tenant. A protocol 4 client's `Nylorun-Tenant` naming
-another Tenant (or malformed), a Tenant that
-could not be opened and rejected credentials → opaque `404` with identical body. A path or method no route serves is `404 Route not found`
+another Tenant (or malformed) and a Tenant that
+could not be opened → opaque `404` with identical body. A path or method no route serves is `404 Route not found`
 once the caller is known.
 
 ## API reference (OpenAPI)

package/dist/adapters/streams/s2.js

@@ -8,10 +8,9 @@
  * - One basin per Tenant basin generation, named by `tenantBasinName` (`streams/basin.ts`):
  *   `<basinPrefix>tn-<ulid>` for a Tenant id `tn_<ulid>`, `…-<g base36>` for generation g.
  * - The basin is created with `createStreamOnAppend` and `createStreamOnRead`, so
- *   `sessions/<id>` and `tenant/control` come into existence on first use. Its
- *   default stream config sets infinite retention, which session streams
- *   keep; the control stream is created with a one-day
- *   age-based retention instead (signals are latency hints, not history).
+ *   `sessions/<id>` come into existence on first use. Its default stream config
+ *   sets infinite retention, which session streams keep. There are no other
+ *   streams: signals between processes go over Postgres (D48).
  * - Bodies are JSON text in string records. Record timestamps are S2 arrival
  *   times.
  * - Appends use `appendRetryPolicy: "noSideEffects"`: the SDK retries only
@@ -27,7 +26,7 @@
  *   explicit `{ infinite: {} }` retention to keep session history; this adapter
  *   sets it as the basin default, and s2-lite honours it (a new stream reports
  *   `infinite`). A `streamConfig` sent with an append that creates the stream
- *   is honoured too (used for the control stream). `deleteOnEmpty`
+ *   is honoured too. `deleteOnEmpty`
  *   defaults to disabled (`minAgeSecs: 0`).
  * - **Stream deletion** is immediate for our purposes: right after `DELETE`,
  *   `checkTail` and reads return 404 `stream_not_found` (new read sessions
@@ -59,9 +58,7 @@
  */
 import { AppendInput, AppendRecord, RangeNotSatisfiableError, S2, S2Error, SeqNumMismatchError, } from "@s2-dev/streamstore";
 import { parseBasin, tenantBasinName, validateBasinPrefix } from "../../streams/basin.js";
-import { CONTROL_STREAM, } from "../../streams/types.js";
-/** Retention of the control signal stream. */
-const SIGNAL_RETENTION_SECS = 24 * 60 * 60;
+import {} from "../../streams/types.js";
 /** Records per unary read while reading history. */
 const HISTORY_PAGE = 1000;
 const RETRY_MIN_MS = 100;
@@ -99,9 +96,6 @@ class S2Streams {
             return AppendRecord.string({ body });
         }), {
             ...(options.matchSeq !== undefined ? { matchSeqNum: options.matchSeq } : {}),
-            ...(stream === CONTROL_STREAM
-                ? { streamConfig: { retentionPolicy: { ageSecs: SIGNAL_RETENTION_SECS } } }
-                : {}),
         });
         const handle = this.handle(tenantId, stream);
         let delay = RETRY_MIN_MS / 4;