@askalf/dario 6.12.23 → 6.12.24

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -510,7 +510,7 @@ export declare function detectDrift(t: TemplateData, installedOverride?: string
510
510
  */
511
511
  export declare const SUPPORTED_CC_RANGE: {
512
512
  readonly min: "1.0.0";
513
- readonly maxTested: "2.1.288";
513
+ readonly maxTested: "2.1.289";
514
514
  };
515
515
  /**
516
516
  * Compare two dotted-numeric version strings. Returns negative if `a<b`,
@@ -1211,7 +1211,7 @@ export function detectDrift(t, installedOverride) {
1211
1211
  */
1212
1212
  export const SUPPORTED_CC_RANGE = {
1213
1213
  min: '1.0.0',
1214
- maxTested: '2.1.288',
1214
+ maxTested: '2.1.289',
1215
1215
  };
1216
1216
  /**
1217
1217
  * Compare two dotted-numeric version strings. Returns negative if `a<b`,
package/docs/docker.md CHANGED
@@ -139,15 +139,22 @@ have to re-run `dario login` each time.
139
139
 
140
140
  ## Healthcheck
141
141
 
142
- The image ships a Docker `HEALTHCHECK` that hits `/health` every 30s. Once an
143
- account is loaded and its OAuth is usable the endpoint returns HTTP 200
144
- `{"status":"ok", ...}`; when OAuth is broken/absent it returns HTTP 503
145
- `{"status":"degraded", ...}`. Use the same endpoint for k8s liveness/readiness
146
- probes:
142
+ dario has two health endpoints, and they answer different questions:
143
+
144
+ - `/livez` is **liveness**: HTTP 200 `{"status":"ok"}` whenever the server is accepting
145
+ requests, whatever the account or OAuth state.
146
+ - `/health` is **readiness**: HTTP 200 `{"status":"ok", ...}` once an account is loaded and
147
+ its OAuth is usable, HTTP 503 `{"status":"degraded", ...}` while OAuth is broken or no
148
+ account exists yet.
149
+
150
+ The image ships a Docker `HEALTHCHECK` that hits `/health` every 30s, so
151
+ `depends_on: service_healthy` waits for a usable account. On Kubernetes, point the
152
+ liveness probe at `/livez`: a broken or missing account is not fixed by restarting the
153
+ pod, and a liveness probe on `/health` would restart it in a loop.
147
154
 
148
155
  ```yaml
149
156
  livenessProbe:
150
- httpGet: { path: /health, port: 3456 }
157
+ httpGet: { path: /livez, port: 3456 }
151
158
  periodSeconds: 30
152
159
  readinessProbe:
153
160
  httpGet: { path: /health, port: 3456 }
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@askalf/dario",
3
- "version": "6.12.23",
3
+ "version": "6.12.24",
4
4
  "description": "Use your Claude and ChatGPT subscriptions in Cursor, Cline, Aider, Claude Code and the Agent SDK — at subscription pricing, not per-token API bills. One local Anthropic + OpenAI-compatible endpoint: either plan answers either wire shape, with automatic failover when one hits its limit.",
5
5
  "type": "module",
6
6
  "bin": {