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.
npx skills add shyftlabs/continuum --skill continuum-llm-providersgit clone --depth 1 https://github.com/shyftlabs/continuumWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/shyftlabs/continuum/continuum-llm-providers)<a href="https://agentmods.dev/skills/shyftlabs/continuum/continuum-llm-providers"><img src="https://agentmods.dev/badge/skills/shyftlabs/continuum/continuum-llm-providers/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/shyftlabs/continuum/continuum-llm-providers"><img src="https://agentmods.dev/badge/skills/shyftlabs/continuum/continuum-llm-providers.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 5.1 | $0.00105 | $0.01971 |
| Opus 5 | $0.00053 | $0.00986 |
| Sonnet 5 | $0.00021 | $0.00394 |
| Haiku 4.5 | $0.00011 | $0.00197 |
Grade A, and why
continuum-llm-providers 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 5d 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.
How it starts
The opening of the file, as written. The whole thing — 264 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Continuum LLM Providers Skill
Authoritative source: docs/llm.md.
⚠️ LiteLLM is removed. All providers are called directly via their official SDKs. Don't suggest
litellm.*imports or any LiteLLM code paths.
Provider routing (by model-string prefix)
When SMART_GATEWAY_URL is not set, routing is by model-string prefix:
| Prefix | Provider | SDK |
|---|---|---|
gemini/..., google/... |
Gemini | OpenAI SDK against Gemini's OpenAI-compat endpoint |
claude/..., anthropic/..., claude-... |
Anthropic | Anthropic SDK |
anything else (gpt-*, azure/..., openai/...) |
OpenAI | OpenAI SDK |
agent.model = "gpt-4o-mini" # OpenAI
agent.model = "claude-sonnet-4-20250514" # Anthropic
agent.model = "gemini/gemini-2.5-flash" # Gemini
Smart Gateway (multi-provider routing)
When SMART_GATEWAY_URL is set, all agents automatically route through
GatewayProvider regardless of the model field. The gateway speaks
OpenAI-compatible API and translates to the upstream provider internally.
SMART_GATEWAY_URL=http://localhost:8787/v1
SMART_GATEWAY_API_KEY=your-key
SMART_GATEWAY_DEFAULT_MODE=modest # strict | modest | quality
No code changes needed — get_provider() detects the env var and switches
automatically.
Per-agent routing mode
Override the default mode on any agent:
agent = BaseAgent(
name="high-quality-agent",
instructions="...",
model="gpt-4o-mini", # sent to gateway as hint
gateway_mode="quality", # overrides SMART_GATEWAY_DEFAULT_MODE
)
gateway_mode |
Gateway tier | Model category |
|---|---|---|
"strict" |
cheap |
ultra-cheap / small models |
"modest" |
mid |
small / mid models (default) |
"quality" |
quality |
frontier models |
How the model field works with gateway
The gateway re-encodes the model as auto/<tier> internally:
"gpt-4o-mini"→"auto/mid"(when mode is modest)- Pass
"openai/gpt-4o-mini"to explicitly pin provider + model - Pass
"auto/quality"to let the gateway choose the best quality model
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.
- 5d ago First seen · 264 lines · 105 tokens per session scan A 0d806d8cb13e
continuum-llm-providers is a skill published in the GitHub repository shyftlabs/continuum (84 stars, last pushed 5d ago), licensed Apache-2.0. It adds 105 tokens to every session and 1,971 once invoked, about $0.0005 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 2026-09-03.
Other skills, from other repositories
spark-environment-setup
Set up a working ML training/inference environment on NVIDIA DGX Spark (GB10, aarch64, CUDA 13). Use when installing PyTorch/Unsloth/TRL/vLLM on DGX Spark, hitting libcudart or wheel-ABI errors on aarch64, or choosing between NGC containers and bare pip installs.
spark-training-gotchas
Preflight and diagnose the ten known failure modes for ML training on NVIDIA DGX Spark. Use when a training run on DGX Spark fails to start, OOMs below the 128GB limit, slows down mid-run, or before any multi-hour training job on GB10.
Vizra ADK Memory System
Implement persistent memory, session context, and vector memory (RAG) for AI agents.
foundations-information-theory
Information-theory primitives for AI systems, entropy, mutual information, KL, compression, channel limits, MDL, bottlenecks, and signal quality. Use when quantifying information.
software-ai-integration
Applies production AI integration patterns for chat, structured output, guardrails, provider routing, and AI UX. Use when adding LLM-powered features to an application.
ai-post-training
Post-training and alignment: reward modeling, RLHF/PPO, DPO/DAAs, GRPO, RLVR, RLAIF, over-optimization. Use when adapting an SFT model with preference or verifiable-reward signals.