What is maintain-dynamic-plugins?

A guide for maintaining NeMo Relay’s dynamically loaded plugins and worker plugins. Dynamic plugins are extensions loaded separately from the main program; worker plugins communicate with NeMo Relay through gRPC, a system for calling services over a network.

Use it when changing Rust plugin loaders, native SDKs, gRPC worker communication, Python worker code, plugin manifests, or related release workflows.

It keeps the boundaries between the core program, native plugins, and worker processes safe and consistent. It also helps keep the SDKs, protocol definitions, documentation, tests, and packaging aligned.

maintain-dynamic-plugins — Skill by NVIDIA

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods

npx agentmods add skills/nvidia/nemo-relay/maintain-dynamic-plugins

Any agent

npx skills add NVIDIA/NeMo-Relay --skill maintain-dynamic-plugins

Clone the repo

git clone --depth 1 https://github.com/NVIDIA/NeMo-Relay

Made for: Claude Code, Codex.

Per session 39 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,080 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5	$0.00039	$0.01080
Opus 5	$0.00019	$0.00540
Sonnet 5	$0.00008	$0.00216
Haiku 4.5	$0.00004	$0.00108

Measured 2d ago against content hash 0333d545953c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

maintain-dynamic-plugins scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.agents/skills/maintain-dynamic-plugins/SKILL.md · 106 lines

How it starts

The opening of the file, as written. The whole thing — 106 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Maintain Dynamic Plugins

Companion Guidance

Use karpathy-guidelines, validate-change, maintain-packaging, and contribute-docs alongside this skill when implementation, packaging, CI, or documentation changes are involved.

Use this skill for plugin.kind = "rust_dynamic", plugin.kind = "worker", nemo-relay-plugin, nemo-relay-worker, nemo-relay-worker-proto, nemo-relay-types, and the Python nemo-relay-plugin package.

Rules

Keep the stable boundary explicit: native plugins cross a C ABI; worker plugins cross grpc-v1.
Do not pass Rust runtime types, trait objects, futures, or allocator-owned strings across the native dynamic-library boundary.
Typed native middleware futures run on the SDK-owned Tokio executor. Keep subscribers synchronous and preserve raw synchronous ABI registrations.
Define closed worker transport structures in protobuf when generated clients must enforce their fields. Keep open application payloads lossless by using JsonValue or JsonEnvelope rather than google.protobuf.Value.
Keep relay-plugin.toml dynamic records separate from generic runtime components. Enabled dynamic records may synthesize internal component specs; disabled records stay inspectable but unloaded.
Relay 0.8 establishes the native API 1 and grpc-v1 canonical ToolExecutionResult baseline. Require every dynamic plugin to rebuild and declare a compat.relay range that excludes versions before 0.8. Recommend >=0.8.0,<1.0; open-ended or narrower 0.8-or-newer ranges are valid.
Treat compat.relay as the plugin author's compatibility assertion, not proof that an artifact was rebuilt. Do not add a legacy raw-result adapter.
Relay 0.8 retains the grpc-v1 identifier and nemo.relay.worker.v1 package while changing the tool-result protobuf types; every worker must regenerate its bindings and rebuild. Native ABI v4 permits append-only host-table extensions guarded by struct_size; existing fields must remain frozen at their original offsets. Incompatible native JSON, reordered or replaced native fields, or incompatible worker protobuf changes must bump native_api or worker_protocol.
Do not add tests under src; Rust tests belong in crate tests/ trees and Python SDK tests belong under python/tests.
Native and worker plugins are trusted extensions. Document that native plugins are in-process and unsandboxed; worker plugins provide process isolation but not a security sandbox.

Read the full file on GitHub · 106 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

2d ago First seen · 106 lines · 39 tokens per session scan A 0333d545953c

Subscribe to this mod's changes

maintain-dynamic-plugins is a skill published in the GitHub repository NVIDIA/NeMo-Relay (132 stars, last pushed 3d ago), licensed Apache-2.0. It adds 39 tokens to every session and 1,080 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Building agents agents ai atif atof claude-code codex

agent-repo-architect

Agent skill for repo-architect - invoke with $agent-repo-architect.

ruvnet/ruflo · 21 tokens

agent-performance-analyzer

Agent skill for performance-analyzer - invoke with $agent-performance-analyzer.

ruvnet/ruflo · 19 tokens

agent-coordination

Agent spawning, lifecycle management, and coordination patterns. Manages 60+ agent types with specialized capabilities. Use when: spawning agents, coordinating multi-agent tasks, managing agent pools. Skip when: single-agent work, no coordination needed.

ruvnet/ruflo · 51 tokens

learn-mcp

Focused interactive tutor for the Model Context Protocol (MCP) path in AI Engineering from Scratch. Start or resume this route when a learner wants to build, secure, debug, verify, or operate MCP clients, servers, transports, gateways, registries, or conformance gates. Teaches one lesson per invocation and records…

rohitg00/ai-engineering-from-scratch · 77 tokens

learn-agent-skills

Focused interactive tutor for the Agent Skills Engineering path in AI Engineering from Scratch. Start or resume this route when a learner wants to create, discover, invoke, secure, evaluate, package, or port Agent Skills. Teaches one lesson per invocation and records evidence in AGENT-SKILLS-LEARNING.md.

rohitg00/ai-engineering-from-scratch · 67 tokens

agent-self-scheduling

Schedule AI agent runs with cron, loops, or external clocks while avoiding unsafe tight autonomous timers.

sickn33/agentic-awesome-skills · 24 tokens