> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.nvidia.com/nemo/relay/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.nvidia.com/nemo/relay/_mcp/server.

# Known Issues

This page lists current limitations and support notes for the release
documentation set.

## NVIDIA NeMo Relay 0.5

These notes apply to the NVIDIA NeMo Relay 0.5 release.

* Go and the raw C FFI surface are experimental and source-first. Generated
  API pages cover Rust, Python, and Node.js; experimental bindings do not have
  the same generated documentation depth.
* The NeMo Relay CLI is experimental. Coding-agent observability support varies
  with host plugin, hook, and provider-routing capabilities. Hooks alone cannot
  produce complete LLM request and response spans.
* Complete first-request capture in Codex plugin mode depends on Codex firing
  an installed hook before the first provider request.
* Node.js 24 or newer is required for Node.js binding and package workflows.
* OpenClaw support uses public hook-backed telemetry with partial security and
  optimization support. Security is limited to pre-tool conditional guardrails,
  and optimization is limited to adaptive telemetry unless the integration owns
  a managed execution path.
* The NeMo Guardrails remote backend depends on the availability, latency, and
  policy behavior of the configured remote service. Its local backend requires
  Python 3.11 or newer and `nemoguardrails==0.22.0`.
* The PII redaction plugin currently provides deterministic local backend
  support. Local-model backend configuration is reserved for future expansion.
* Pricing estimates depend on configured model-pricing sources and their
  freshness. Unknown model pricing and missing token data leave cost absent
  instead of defaulting to zero.
* ATOF streaming endpoints and remote ATIF storage depend on reachable,
  correctly configured destinations. Endpoint failures do not block local file
  output or other configured endpoints.
* Native dynamic plugins run in-process and are not sandboxed. `grpc-v1` worker
  plugins isolate a process but do not provide a security sandbox.
* `LLMRequest` objects in the Python binding should be treated as immutable.
  Request middleware that changes content should return a new request object.
* Native subscriber callbacks are delivered asynchronously. Flush subscribers
  before relying on callback side effects, captured event lists, files, or
  exporter output. Deregistering a subscriber affects future emissions, but
  callbacks from already-queued event snapshots can still run.

### Compatibility and Migration Notes

* Move plugin configuration from `config.toml` to `plugins.toml`. The
  `--plugin-config` option is unavailable in 0.5.
* Registered LLM request and tool execution intercepts must return their
  canonical outcome types. Rebuild development native plugins and `grpc-v1`
  workers against 0.5.
* The experimental `nemo-relay-wasm` package is no longer built, published, or
  supported by this repository.
* The `nemo-relay cursor` and `hook-forward cursor` entry points are no longer
  available. Patch-based integrations are also no longer maintained.

### Fixed in NVIDIA NeMo Relay 0.5

* Gateway shutdown flushes queued events before it exits.
* Dynamic worker invocations support cancellation, and the CLI manages Python
  dynamic-plugin environments.
* ATOF file export creates a missing output directory before opening JSONL
  output.
* CLI status output reports daemon binding, exporter status, and remote ATIF
  storage destinations more reliably.
* Codex host-plugin discovery uses the supported text output from the Codex
  plugin command.

### Fixed in Prior Releases

#### NVIDIA NeMo Relay 0.4

* ATIF shutdown no longer deadlocks queued subscribers.
* Sanitized LLM requests are resolved from annotations for observability output.
* Structured ATIF tool results and Hermes tool-result observations are
  preserved more reliably.
* Hermes routed-provider spans, wrapped ATIF fidelity, subagent lineage, and
  error-path export consistency are covered and corrected.
* OpenClaw observability output is more consistent for nested subagents, model
  timing diagnostics, hook-backed provenance, placeholder replay, and
  hook-only fallback exports.
* LangChain serialization handles wrapped integration payloads more reliably.
* Plugin context-manager teardown avoids the previous deadlock path.
* Node.js `withScope` callbacks receive a real `ScopeHandle`.
* Deep Agents model responses are annotated for downstream observability.

#### NVIDIA NeMo Relay 0.3

* Managed LLM start events are emitted before execution intercepts.
* Coding-agent trace scopes are aligned with NeMo Relay agent scope semantics.
* ATIF tool observations are correlated with their matching tool calls.
* OpenClaw tool call replay visibility is preserved.
* LangChain serialization handles wrapped integration payloads more reliably.

#### Earlier Releases

* Enabled TLS support for OTLP HTTP export.
* Preserved Go scope stacks across OS threads.