Troubleshooting
Start Here
Section titled “Start Here”aegis statusaegis logs coreaegis diagnostics deployment-readinessaegis diagnostics native-readinessCommon Problems
Section titled “Common Problems”| Symptom | Likely cause | First check |
|---|---|---|
| Core is unreachable | Core not started, wrong profile, port conflict | aegis status |
| Chat cannot plan tools | Provider missing or invalid | provider environment keys and provider status |
| Worker says it cannot act | Local Gateway not connected or tools not advertised | aegis worker tools --no-exec |
| Mobile app cannot connect | Wrong Core URL, no LAN/private/tunnel route, token not paired | Core URL and pairing state |
| Channel receives messages but does not act | Handler should route ordinary prose to Manager, not parse keywords | gateway logs and provider trace |
| Native app shows missing permissions | OS-level permission not granted | platform Settings / privacy permissions |
Mobile Pairing
Section titled “Mobile Pairing”Mobile devices use scoped native node tokens. They should not store the long-lived owner token.
If pairing fails:
- Confirm Core URL is reachable from the phone.
- Confirm the device is on the same trusted network, VPN, Tailscale, or tunnel.
- Generate a fresh setup link or QR code from the desktop app.
- Approve the device in Core.
- Revoke stale device tokens if needed.
When to Use Logs
Section titled “When to Use Logs”Use logs to inspect runtime behavior, not to bypass product state:
aegis logs coreFor AI tool-use issues, prefer provider trace and tool ledger evidence over guesses from raw logs.