v2.0 to v2.0

12 added, 12 removed. Audit A to A.

---

name: provider-patterns

description: Configure per-provider behavior, understand streaming quirks, and use the 7-hook adapter system for optimal performance across LLM providers.

compatibility: Reactive Agents TypeScript projects using @reactive-agents/*

metadata:

author: reactive-agents

version: "2.0"

tier: "capability"

---

# Provider Patterns

## Agent objective

Produce a builder with the correct provider + model + any provider-specific configuration; know which providers need special handling for streaming and tool calls.

## When to load this skill

- Configuring a non-Anthropic provider

- Debugging tool call or streaming issues specific to one provider

- Using local models (Ollama) or proxy routing (LiteLLM)

- Enabling extended thinking or provider-specific model options

## Implementation baseline

```ts

// Anthropic — highest quality, native FC, prompt caching

const agent = await ReactiveAgents.create()

.withProvider("anthropic")

- .withModel("claude-opus-4-6")

+ .withModel("claude-sonnet-4-6")

.withReasoning({ defaultStrategy: "adaptive" })

.withTools()

.build();

// Local Ollama model

const agent = await ReactiveAgents.create()

.withProvider("ollama")

.withModel("qwen2.5:7b")

.withReasoning({ defaultStrategy: "reactive", maxIterations: 6 })

.withTools({ allowedTools: ["web-search"] })

.build();

```

## Provider selection guide

| Provider | Best for | Key notes |

|----------|----------|-----------|

| `"anthropic"` | Production, highest quality | Native FC, prompt caching, streaming |

| `"openai"` | GPT-4o, broad compatibility | Native FC, streaming |

| `"gemini"` | Multimodal, long context | Native FC; `functionResponse.name` quirk |

| `"ollama"` | Local, privacy-first | Tool calls arrive on `chunk.done` |

| `"litellm"` | Proxy routing, cost optimization | OpenAI-compatible; use for Groq, OpenRouter, etc. |

| `"test"` | Unit tests, CI | Returns deterministic mock responses |

## Key patterns

### Extended thinking (Anthropic)

```ts

.withProvider("anthropic")

.withModel({ model: "claude-opus-4-6", thinking: true })

// Enables extended thinking — model reasons before responding

// Higher quality on complex reasoning tasks; adds latency and cost

```

### LiteLLM for provider routing

```ts

// Groq, OpenRouter, Bedrock, Vertex — all through LiteLLM

.withProvider("litellm")

.withModel("groq/llama-3.1-70b-versatile")

// Model name format: "provider/model-name" as per LiteLLM docs

```

### Circuit breaker for unreliable providers

```ts

.withProvider("ollama")

.withModel("llama3:8b")

.withCircuitBreaker({

failureThreshold: 3, // open after 3 consecutive failures

cooldownMs: 30_000, // wait 30s before half-open probe

halfOpenRequests: 1,

})

.withRateLimiting({ requestsPerMinute: 10 })

```

### Enabling temperature and sampling

```ts

.withModel({ model: "gpt-4o", temperature: 0.2 }) // more deterministic

.withModel({ model: "claude-sonnet-4-6", temperature: 0.9 }) // more creative

```

## 7 adapter hooks (automatic — no configuration needed)

- These hooks run automatically and adapt prompts/behavior for each provider:

+ These hooks run automatically and adapt prompts/behavior for each provider's strengths:

- | Hook | What it does |

- |------|-------------|

- | `taskFraming` | Wraps task in provider-optimal framing |

- | `toolGuidance` | Injects provider-specific tool-use instructions |

- | `continuationHint` | Tells model to continue after tool results |

- | `errorRecovery` | Recovery prompt on tool errors |

- | `synthesisPrompt` | Final answer synthesis guidance |

- | `qualityCheck` | Post-step quality assessment |

- | `systemPromptPatch` | Provider-specific system prompt additions |

+ | Hook | Method name | What it does |

+ |------|-------------|-------------|

+ | `taskFraming` | `adapter.frameTask(task, tier)` | Wraps task in provider-optimal framing |

+ | `toolGuidance` | `adapter.injectToolGuidance(prompt)` | Injects provider-specific tool-use instructions |

+ | `continuationHint` | `adapter.getContinuationHint(lastMsg)` | Tells model to continue after tool results |

+ | `errorRecovery` | `adapter.getErrorRecoveryPrompt(error)` | Recovery prompt on tool errors |

+ | `synthesisPrompt` | `adapter.getSynthesisPrompt(context)` | Final answer synthesis guidance |

+ | `qualityCheck` | `adapter.checkQuality(output, task)` | Post-step quality assessment |

+ | `systemPromptPatch` | `adapter.patchSystemPrompt(base)` | Provider-specific system prompt additions |

- Adapter selection is automatic via `selectAdapter(capabilities, tier)`.

+ Adapter selection is automatic via `selectAdapter(capabilities, tier)`. Each provider (Anthropic, OpenAI, Gemini, Ollama) has an adapter with specialized implementations.

## Builder API reference

| Method | Key params | Notes |

|--------|-----------|-------|

| `.withModel(m)` | `string \| { model, thinking?, temperature? }` | `thinking: true` = extended reasoning |

| `.withCircuitBreaker(cfg?)` | `{ failureThreshold?, cooldownMs?, halfOpenRequests? }` | Auto-retry with backoff |

| `.withRateLimiting(cfg)` | `{ requestsPerMinute?, tokensPerMinute?, maxConcurrent? }` | |

## Pitfalls

- `"groq"` and `"openrouter"` are **not** valid provider names — use `"litellm"` with the appropriate model prefix

- **Gemini**: `functionResponse.name` must use `msg.toolName`, not hard-coded `"tool"` — framework handles this but custom tool parsers must follow the same pattern

- **Ollama**: tool_calls arrive on `chunk.done`, not during the stream — don't parse mid-stream chunks for tool calls

- **Anthropic streaming**: use raw `streamEvent`, not helper events (`inputJson` fires before `contentBlock` in streaming FC)

- `thinking: true` requires a model that supports extended thinking — verify model capability before enabling

- LiteLLM model names are `"provider/model"` format — check LiteLLM docs for exact names