AWS provider not working with more than a single secret (batch mode)

[lokkju]

when using the AWS provider, if I request secrets one at a time (fnox get) they all work fine; but fnox exec gives me warnings for all except the first secret listed:

fnox exec -- python -c "import os; print(os.environ);"|grep XXXX
2026-01-10T03:19:19.090396Z  WARN fnox::secret_resolver: Error resolving secret 'XXXXXX2': Provider error: AWS Secrets Manager batch call failed: service error
2026-01-10T03:19:19.090413Z  WARN fnox::secret_resolver: Error resolving secret 'XXXXXX3': Provider error: AWS Secrets Manager batch call failed: service error
2026-01-10T03:19:19.090415Z  WARN fnox::secret_resolver: Error resolving secret 'XXXXXX4': Provider error: AWS Secrets Manager batch call failed: service error

fnox doctor:

🏥 Fnox Doctor Report
====================

📄 Configuration:
  File: fnox.toml
  Profile: default
  Status: ✓ Loaded successfully

🔐 Secrets:
  Count: 4
  With value source: 4
  Required: 0
  Using providers: 4
  Secrets:
    ✓ XXXXXX1 (provider: aws)
    ✓ XXXXXX2 (provider: aws)
    ✓ XXXXXX3 (provider: aws)
    ✓ XXXXXX4 (provider: aws)

🔧 Providers:
  Count: 1
  Providers:
    aws (aws-sm)

🌍 Environment:
  FNOX_PROFILE: (not set)

🔍 Provider Health:
  aws: Testing... ✓

📊 Summary:
  Total secrets: 4
  Total providers: 1

💡 Tips:
  - Run 'fnox check' to validate your configuration

fnox check

Checking configuration for profile: default
Found 4 secret(s) in profile
Found 1 provider(s) in profile
✓ Configuration is healthy

fnox -V:

fnox 1.7.0

[lokkju]

Found the problem: in my configuration, I had the same value set twice for the same provider; this was causing the AWS batch request to fail. I'm not sure if this should be considered a bug, or if it's just something to work around.

Basically:

SERVICE_ONE_KEY = { "provider" = "aws", value = "the_master_service_key" }
SERVICE_TWO_KEY = { "provider" = "aws", value = "the_master_service_key" }

Once I removed the duplicate, it all works fine