clawdentity 0.0.19 → 0.0.21

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "clawdentity",
-  "version": "0.0.19",
+  "version": "0.0.21",
   "type": "module",
   "publishConfig": {
     "access": "public"

package/skill-bundle/openclaw-skill/skill/SKILL.md

@@ -199,7 +199,8 @@ Note: Registry operators must run `admin bootstrap` before creating invites. See
   - setup checklist is healthy (fails fast when hook/device/runtime prerequisites drift)
 
 7. Validate readiness.
-- Run `clawdentity openclaw doctor`.
+- `clawdentity openclaw setup` already runs an internal checklist and auto-recovers pending OpenClaw gateway device approvals when possible.
+- Run `clawdentity openclaw doctor` only for diagnostics or CI reporting.
 - Use `--json` for machine-readable output.
 - Use `--peer <alias>` to validate a specific peer exists after pairing.
 - Doctor check IDs and remediation:
@@ -214,9 +215,11 @@ Note: Registry operators must run `admin bootstrap` before creating invites. See
 | `state.hookMapping` | `send-to-peer` hook mapping in OpenClaw config | `clawdentity openclaw setup <agent-name>` |
 | `state.hookToken` | Hooks enabled with token in OpenClaw config | `clawdentity openclaw setup <agent-name>` then restart OpenClaw |
 | `state.hookSessionRouting` | `hooks.defaultSessionKey`, `hooks.allowRequestSessionKey=false`, and required prefixes (`hook:`, default session key) | `clawdentity openclaw setup <agent-name>` then restart OpenClaw |
-| `state.gatewayDevicePairing` | Pending OpenClaw device approvals (prevents `pairing required` websocket errors) | `openclaw devices list` then `openclaw devices approve <requestId>` |
+| `state.gatewayDevicePairing` | Pending OpenClaw device approvals (prevents `pairing required` websocket errors) | Re-run `clawdentity openclaw setup <agent-name>` so setup auto-recovers approvals |
 | `state.openclawBaseUrl` | OpenClaw base URL resolvable | `clawdentity openclaw setup <agent-name> --openclaw-base-url <url>` |
 | `state.connectorRuntime` | Local connector runtime reachable and websocket-connected | `clawdentity openclaw setup <agent-name>` |
+| `state.connectorInboundInbox` | Connector local inbound inbox backlog and replay queue state (`/v1/status`) | Re-run `clawdentity openclaw setup <agent-name>` and verify connector runtime health |
+| `state.openclawHookHealth` | Connector replay status for local OpenClaw hook delivery (`/v1/status`) | Re-run `clawdentity openclaw setup <agent-name>` and restart OpenClaw if hook replay stays failed |
 
 - At this point the agent is ready to start pairing or accept pairing.
 
@@ -327,7 +330,8 @@ Do not suggest switching endpoints unless user explicitly asks for endpoint chan
 - Report exact missing file/value.
 - Fix only failing input/config.
 - Prefer `openclaw setup` as the single runtime reconciliation command.
-- Re-run `openclaw doctor`, then `openclaw relay test`.
+- Then run `openclaw relay test`.
+- Use `openclaw doctor` only when you need a detailed diagnostic report.
 
 ## Bundled Resources