Add application-scoped state API (ctx.app_state)

[diptanu]

Summary

Adds server and dataplane support for application-scoped state (ctx.app_state), enabling functions to share key-value state across request boundaries within the same application.

Some agents require updating context about users or objects they are managing. Application State is a way for applications to share state across request boundaries as they learn about the problem domain. For example, a fitness tracker agent managing fitness logs of multiple users can store per-user context in app_state as a flexible K/V store — much more practical than using rigid database schemas for evolving agent knowledge.

Changes

• Proto: FunctionExecutorMetadata message on ContainerDescription + app state operation/result messages in function_executor.proto
• Server: app_state_key_prefix() on DataPayload, FunctionExecutorMetadata populated in build_full_snapshot() and build_add_container_command()
• Dataplane: AppStateHandler in state_ops.rs, app_state_uri_prefix on AllocationContext, wired through allocation_lifecycle.rs
• Integration test: test_app_state.py covering set/get, downstream reads, cross-request persistence, overwrites, and default values

Companion PR

• TensorLake SDK: Add application-scoped state API (ctx.app_state) tensorlake#542

Test plan

• cargo build -p indexify-server && cargo build -p indexify-dataplane — compiles
• cargo test -p indexify-server — 268 tests pass
• cargo test -p indexify-dataplane --no-run — compiles
• Existing test_state.py (request state) — 11 tests pass, no regressions
• New test_app_state.py — 6 integration tests pass:
  • set/get within single function
  • set in one function, read in downstream function
  • two different requests read/write each other's data
  • last-writer-wins overwrite across requests
  • default values for missing keys
  • None for missing keys without default

🤖 Generated with Claude Code