[Provider Bug] Provider crashes with nil pointer panic in serialBroadcaster.syncSequence caused by ghost leases

[drape76]

Describe the bug
Provider crashes with a nil pointer panic in serialBroadcaster.syncSequence (tx.go:541). The crash is triggered by spam tenants creating ghost leases — leases that are closed on-chain but remain in the provider's memory as active deployments. After repeated failed deployment health checks (~10+ attempts with "available replicas below target"), the provider panics and exits with code 1 or 2.

account address = akash1chnhnu50f6hv98xl0m7xm95vel457ysp32uwpj
deployment sequence = 25857450 (example, recurs with new ghost leases)
group sequence = 1

To Reproduce
Spam tenant creates a lease with the provider
Tenant closes deployment on-chain but provider retains it in memory
Provider deployment-monitor repeatedly checks service replicas, gets available=0 for 10+ attempts
Provider panics with nil pointer dereference in serialBroadcaster.syncSequence

Expected behavior
Provider should handle closed/invalid leases gracefully without panicking. Ghost leases should be detected and cleaned up without requiring a full provider restart.

Crash log
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x28 pc=0x5e8fffd]

goroutine 56 [running]:
pkg.akt.dev/go/node/client/v1beta3.(*serialBroadcaster).syncSequence(...)
pkg.akt.dev/go@v0.1.9/node/client/v1beta3/tx.go:541 +0xfd
pkg.akt.dev/go/node/client/v1beta3.(*serialBroadcaster).broadcaster(...)
pkg.akt.dev/go@v0.1.9/node/client/v1beta3/tx.go:572 +0xa0c

Additional context
Provider version: 0.10.7 (latest)
Helm chart: 14.0.8
Kubernetes: K3s
Hardware: Dell PowerEdge R640, 80 vCPU, 384GB RAM
The crash recurs every few hours as new ghost leases appear

Workaround:
automated script that closes ghost leases on-chain hourly and restarts provider, but the underlying panic is not resolved

[drape76]

Relevant logs leading up to crash:
2:30PM DBG service available replicas below target available=0 cmp=deployment-monitor lease=akash1ajpr854uqetp8cvgqcntmweehf2f4ut3xxxfw8/25857450/1/1/akash1chnhnu50f6hv98xl0m7xm95vel457ysp32uwpj manifest-group=akash module=provider-cluster service=app target=1
2:30PM INF check result attempt=9 cmp=deployment-monitor lease=akash1ajpr854uqetp8cvgqcntmweehf2f4ut3xxxfw8/25857450/1/1/akash1chnhnu50f6hv98xl0m7xm95vel457ysp32uwpj manifest-group=akash module=provider-cluster ok=false
2:30PM DBG service available replicas below target available=0 cmp=deployment-monitor lease=akash1ajpr854uqetp8cvgqcntmweehf2f4ut3xxxfw8/25857450/1/1/akash1chnhnu50f6hv98xl0m7xm95vel457ysp32uwpj manifest-group=akash module=provider-cluster service=app target=1
2:30PM INF check result attempt=10 cmp=deployment-monitor lease=akash1ajpr854uqetp8cvgqcntmweehf2f4ut3xxxfw8/25857450/1/1/akash1chnhnu50f6hv98xl0m7xm95vel457ysp32uwpj manifest-group=akash module=provider-cluster ok=false
2:30PM DBG service available replicas below target available=0 cmp=deployment-monitor lease=akash1ajpr854uqetp8cvgqcntmweehf2f4ut3xxxfw8/25857450/1/1/akash1chnhnu50f6hv98xl0m7xm95vel457ysp32uwpj manifest-group=akash module=provider-cluster service=app target=1
2:30PM INF check result attempt=11 cmp=deployment-monitor lease=akash1ajpr854uqetp8cvgqcntmweehf2f4ut3xxxfw8/25857450/1/1/akash1chnhnu50f6hv98xl0m7xm95vel457ysp32uwpj manifest-group=akash module=provider-cluster ok=false

Panic:
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x28 pc=0x5e8fffd]

goroutine 56 [running]:
pkg.akt.dev/go/node/client/v1beta3.(*serialBroadcaster).syncSequence(, {{0x8d7cdb0, 0xc001b0b220}, {0x8d3ba98, 0xc001732880}, {0x8d0d408, 0xc8120e0}, 0x1163f9, 0x17e5, 0x0, ...}, ...)
pkg.akt.dev/go@v0.1.9/node/client/v1beta3/tx.go:541 +0xfd
pkg.akt.dev/go/node/client/v1beta3.(*serialBroadcaster).broadcaster(, {{0x8d7cdb0, 0xc001b0b220}, {0x8d3ba98, 0xc001732880}, {0x8d0d408, 0xc8120e0}, 0x1163f9, 0x17e5, 0x0, ...})
pkg.akt.dev/go@v0.1.9/node/client/v1beta3/tx.go:572 +0xa0c
created by pkg.akt.dev/go/node/client/v1beta3.newSerialTx in goroutine 1
pkg.akt.dev/go@v0.1.9/node/client/v1beta3/tx.go:314 +0xa53
Exit code: 2

Note:
Lease akash1ajpr854uqetp8cvgqcntmweehf2f4ut3xxxfw8/25857450 was already closed on-chain (reason: lease_closed_invalid) but remained active in provider memory, triggering the deployment-monitor loop that ultimately caused the panic.