refactor(sdk): hard-rename provider proxy client

apps/web/src/content/docs/docs/next/graders/script-graders.mdx

@@ -288,41 +288,41 @@ Prefer Vitest verifiers when the checks naturally fit `expect(...)`. Use `define
 
 **SDK exports:** `defineScriptGrader`, `defineVitestWorkspaceGrader`, `defineWorkspaceGrader`, `Message`, `ToolCall`, `Trace`, `TraceSummary`, `ScriptGraderInput`, `ScriptGraderResult`, `ScriptGraderCheck`, `Workspace`, `WorkspaceCheck`
 
-## Target Access
+## Provider Access
 
-Script graders can call an LLM through a target proxy for metrics that require multiple LLM calls (contextual precision, semantic similarity, etc.).
+Script graders can call an LLM through a provider proxy for metrics that require multiple LLM calls (contextual precision, semantic similarity, etc.).
 
 ### Configuration
 
-Add a `target` block to the grader config:
+Add a `provider` block to the grader config:
 
 ```yaml
 assert:
   - name: contextual-precision
     type: script
     command: [bun, scripts/contextual-precision.ts]
-    target:
+    provider:
       max_calls: 10  # Default: 50
 ```
 
 ### Usage
 
-Use `createTargetClient` from the SDK:
+Use `createProviderClient` from the SDK:
 
 ```typescript
 #!/usr/bin/env bun
-import { createTargetClient, defineScriptGrader } from 'agentv';
+import { createProviderClient, defineScriptGrader } from 'agentv';
 
 export default defineScriptGrader(async ({ input, output }) => {
   const inputText = input
     .filter((message) => message.role === 'user')
     .map((message) => typeof message.content === 'string' ? message.content : '')
     .join('\n');
   const outputText = output ?? '';
-  const target = createTargetClient();
-  if (!target) return { pass: false, score: 0, reason: 'Target not configured' };
+  const provider = createProviderClient();
+  if (!provider) return { pass: false, score: 0, reason: 'Provider proxy not configured' };
 
-  const response = await target.invoke({
+  const response = await provider.invoke({
     question: `Is this relevant to: ${inputText}? Response: ${outputText}`,
     systemPrompt: 'Respond with JSON: { "relevant": true/false }'
   });
@@ -336,14 +336,14 @@ export default defineScriptGrader(async ({ input, output }) => {
 });
 ```
 
-Use `target.invokeBatch(requests)` for multiple calls in parallel.
+Use `provider.invokeBatch(requests)` for multiple calls in parallel.
 
-**Environment variables** (set automatically when `target` is configured):
+**Environment variables** (set automatically when `provider` is configured):
 
 | Variable | Description |
 |----------|-------------|
-| `AGENTV_TARGET_PROXY_URL` | Local proxy URL |
-| `AGENTV_TARGET_PROXY_TOKEN` | Bearer token for authentication |
+| `AGENTV_PROVIDER_PROXY_URL` | Local proxy URL |
+| `AGENTV_PROVIDER_PROXY_TOKEN` | Bearer token for authentication |
 
 ## Advanced Input Fields
 
@@ -353,7 +353,7 @@ Beyond the basic fields (`input`, `output`, `expected_output`), script graders r
 |-------|------|-------------|
 | `input` | `Message[]` | Full resolved input message array |
 | `output` | `string \| null` | Final answer / scored result only |
-| `messages` | `Message[]` | Transcript messages from the target execution |
+| `messages` | `Message[]` | Transcript messages from the provider execution |
 | `expected_output` | `Message[]` | Expected/reference output messages |
 | `output_path` | `string` | Temp file containing large final answer JSON, when `output` is omitted |
 | `input_files` | `string[]` | Paths to input files referenced in the eval |

examples/README.md

@@ -47,10 +47,10 @@ Focused demonstrations of specific AgentV capabilities. Each example includes it
 - [assert-set](features/assert-set/) - Assertion grouping patterns
 - [weighted-graders](features/weighted-graders/) - Weighted graders
 - [execution-metrics](features/execution-metrics/) - Metrics tracking (tokens, cost, latency)
-- [script-grader-with-llm-calls](features/script-grader-with-llm-calls/) - script graders with target proxy for LLM calls
+- [script-grader-with-llm-calls](features/script-grader-with-llm-calls/) - script graders with provider proxy for LLM calls
 - [batch-cli](features/batch-cli/) - Batch CLI evaluation
 - [document-extraction](features/document-extraction/) - Document data extraction
-- [local-cli](features/local-cli/) - Local CLI targets
+- [local-cli](features/local-cli/) - Local CLI providers
 - [compare](features/compare/) - Baseline comparison
 - [deterministic-graders](features/deterministic-graders/) - Deterministic assertions (contains, regex, JSON validation)
 - [vitest-workspace-grader](features/vitest-workspace-grader/) - Vitest-style deterministic workspace verifiers
@@ -92,7 +92,7 @@ example-name/
 │   └── *.md                  # LLM grader prompts (optional)
 ├── scripts/                  # Helper scripts (optional)
 ├── .agentv/
-│   └── providers.yaml          # Target configuration (optional)
+│   └── providers.yaml          # Provider configuration (optional)
 ├── package.json              # Dependencies (if using agentv)
 └── README.md                 # Example documentation
 ```

examples/features/README.md

@@ -40,7 +40,7 @@ Focused examples for specific AgentV capabilities. Find your use case below, the
 | Example | Description |
 |---------|-------------|
 | [script-grader-sdk](script-grader-sdk/) | TypeScript script graders using `defineScriptGrader()` from `agentv` |
-| [script-grader-with-llm-calls](script-grader-with-llm-calls/) | script graders that make LLM calls via a target proxy |
+| [script-grader-with-llm-calls](script-grader-with-llm-calls/) | script graders that make LLM calls via a provider proxy |
 | [eval-assert-demo](eval-assert-demo/) | script graders runnable both in a suite and individually via `agentv eval assert` |
 | [functional-grading](functional-grading/) | Install dependencies, compile, and run tests against agent-generated code |
 
@@ -51,7 +51,7 @@ Focused examples for specific AgentV capabilities. Find your use case below, the
 |---------|-------------|
 | [trajectory-assertions-simple](trajectory-assertions-simple/) | Validate expected tool calls with Promptfoo trajectory assertions |
 | [trajectory-assertions-advanced](trajectory-assertions-advanced/) | Promptfoo trajectory assertions with `expected_output` and per-call checks |
-| [latency-assertions](latency-assertions/) | Tool sequence and argument checks for a latency-flavored mock target |
+| [latency-assertions](latency-assertions/) | Tool sequence and argument checks for a latency-flavored mock provider |
 | [tool-evaluation-plugins](tool-evaluation-plugins/) | F1 precision/recall scoring for tool-call accuracy |
 | [trace-evaluation](trace-evaluation/) | Inspect agent internals: LLM call counts, tool executions, step durations |
 
@@ -103,10 +103,10 @@ Focused examples for specific AgentV capabilities. Find your use case below, the
 | [workspace-setup-script](workspace-setup-script/) | Multi-step setup with a `beforeAll` lifecycle extension |
 | [workspace-multi-repo](workspace-multi-repo/) | Multi-repo environment using a VS Code `.code-workspace` file |
 | [workspace-shared-config](workspace-shared-config/) | Define an environment recipe once and reference it across eval files |
-| [repo-lifecycle](repo-lifecycle/) | Clone a git repo into the workspace and target the agent at it |
+| [repo-lifecycle](repo-lifecycle/) | Clone a git repo into the workspace and point the agent at it |
 | [file-changes](file-changes/) | Capture workspace file changes made by the agent across test runs |
 | [file-changes-graders](file-changes-graders/) | Grade file diffs with rubrics and LLM graders |
-| [local-cli](local-cli/) | Define and invoke local CLI targets |
+| [local-cli](local-cli/) | Define and invoke local CLI providers |
 | [batch-cli](batch-cli/) | Run bulk evaluations from the CLI |
 
 ---
@@ -138,7 +138,7 @@ Focused examples for specific AgentV capabilities. Find your use case below, the
 | [assert-extended](assert-extended/) | Deterministic assertions |
 | [basic](basic/) | Getting started |
 | [basic-jsonl](basic-jsonl/) | Getting started |
-| [batch-cli](batch-cli/) | Workspace & targets |
+| [batch-cli](batch-cli/) | Workspace & providers |
 | [benchmark-tooling](benchmark-tooling/) | Benchmarking |
 | [script-grader-sdk](script-grader-sdk/) | Custom graders |
 | [script-grader-with-llm-calls](script-grader-with-llm-calls/) | Custom graders |
@@ -152,17 +152,17 @@ Focused examples for specific AgentV capabilities. Find your use case below, the
 | [eval-assert-demo](eval-assert-demo/) | Custom graders |
 | [execution-metrics](execution-metrics/) | Cost, latency & tokens |
 | [external-datasets](external-datasets/) | Dataset & input |
-| [file-changes](file-changes/) | Workspace & targets |
-| [file-changes-graders](file-changes-graders/) | Workspace & targets |
+| [file-changes](file-changes/) | Workspace & providers |
+| [file-changes-graders](file-changes-graders/) | Workspace & providers |
 | [functional-grading](functional-grading/) | Custom graders |
 | [input-files-shorthand](input-files-shorthand/) | Dataset & input |
 | [latency-assertions](latency-assertions/) | Tool & agent evaluation |
-| [local-cli](local-cli/) | Workspace & targets |
+| [local-cli](local-cli/) | Workspace & providers |
 | [multi-turn-conversation](multi-turn-conversation/) | LLM grading |
 | [nlp-metrics](nlp-metrics/) | Deterministic assertions |
 | [file-transforms](file-transforms/) | LLM grading |
 | [prompt-template-sdk](prompt-template-sdk/) | TypeScript SDK |
-| [repo-lifecycle](repo-lifecycle/) | Workspace & targets |
+| [repo-lifecycle](repo-lifecycle/) | Workspace & providers |
 | [rubric](rubric/) | LLM grading |
 | [scenarios](scenarios/) | Dataset & prompt templates |
 | [sdk-config-file](sdk-config-file/) | TypeScript SDK |
@@ -181,6 +181,6 @@ Focused examples for specific AgentV capabilities. Find your use case below, the
 | [trial-output-consistency](trial-output-consistency/) | Benchmarking |
 | [trials](trials/) | Benchmarking |
 | [weighted-graders](weighted-graders/) | LLM grading |
-| [workspace-multi-repo](workspace-multi-repo/) | Workspace & targets |
-| [workspace-setup-script](workspace-setup-script/) | Workspace & targets |
-| [workspace-shared-config](workspace-shared-config/) | Workspace & targets |
+| [workspace-multi-repo](workspace-multi-repo/) | Workspace & providers |
+| [workspace-setup-script](workspace-setup-script/) | Workspace & providers |
+| [workspace-shared-config](workspace-shared-config/) | Workspace & providers |

examples/features/script-grader-with-llm-calls/README.md

@@ -141,37 +141,37 @@ The current `extractRetrievalContext()` in `utils.ts` flattens for simplicity. F
 
 ## Security
 
-The target proxy is designed with security in mind:
+The provider proxy is designed with security in mind:
 - Binds to **loopback only** (127.0.0.1) - not accessible from network
 - Uses **bearer token authentication** - unique per execution
 - Enforces **max_calls limit** - prevents runaway costs
 - **Auto-shutdown** - proxy terminates when grader completes
 
 ## Configuration
 
-Enable target access by adding a `target` block to your `script_grader` grader:
+Enable provider access by adding a `provider` block to your `script_grader` grader:
 
 ```yaml
 graders:
   - name: contextual_precision
     type: script
     command: [bun, run, scripts/contextual-precision.ts]
-    target:
+    provider:
       max_calls: 10  # At least N nodes to evaluate
   - name: contextual_recall
     type: script
     command: [bun, run, scripts/contextual-recall.ts]
-    target:
+    provider:
       max_calls: 15  # 1 for extraction + N statements for attribution
 ```
 
 ## Usage in Code
 
 ```typescript
-import { createTargetClient, defineScriptGrader } from 'agentv';
+import { createProviderClient, defineScriptGrader } from 'agentv';
 
 export default defineScriptGrader(async ({ question, config }) => {
-  const target = createTargetClient();
+  const provider = createProviderClient();
   const retrievalContext = config?.retrieval_context ?? [];
 
   // Batch evaluation of all nodes
@@ -180,50 +180,50 @@ export default defineScriptGrader(async ({ question, config }) => {
     systemPrompt: 'Respond with JSON: { "relevant": true/false }'
   }));
 
-  const responses = await target.invokeBatch(requests);
+  const responses = await provider.invokeBatch(requests);
 
   // Calculate weighted precision score...
 });
 ```
 
 ## Querying Proxy Info
 
-You can query information about the target proxy:
+You can query information about the provider proxy:
 
 ```typescript
-const info = await target.getInfo();
-console.log(`Target: ${info.targetName}`);
+const info = await provider.getInfo();
+console.log(`Provider: ${info.providerLabel}`);
 console.log(`Calls: ${info.callCount}/${info.maxCalls}`);
-console.log(`Available targets: ${info.availableTargets.join(', ')}`);
+console.log(`Available providers: ${info.availableProviderLabels.join(', ')}`);
 ```
 
-## Target Override
+## Provider Override
 
-Use different targets for different purposes within the same grader:
+Use different providers for different purposes within the same grader:
 
 ```typescript
 // Use a coding agent for complex tasks
-const agentResponses = await target.invokeBatch(
+const agentResponses = await provider.invokeBatch(
   nodes.map(node => ({
     question: `Is this relevant? ${node}`,
-    target: 'pi'  // Override default target
+    provider: 'pi'  // Override default provider
   }))
 );
 
 // Use a base LLM for simple evaluation
-const response = await target.invoke({
+const response = await provider.invoke({
   question: complexAnalysisPrompt,
-  target: 'gemini-llm'  // Use different target
+  provider: 'gemini-llm'  // Use different provider
 });
 ```
 
 ## Environment Variables
 
-When `target` is configured, these environment variables are automatically set:
-- `AGENTV_TARGET_PROXY_URL` - Local proxy URL (e.g., `http://127.0.0.1:45123`)
-- `AGENTV_TARGET_PROXY_TOKEN` - Bearer token for authentication
+When `provider` is configured, these environment variables are automatically set:
+- `AGENTV_PROVIDER_PROXY_URL` - Local proxy URL (e.g., `http://127.0.0.1:45123`)
+- `AGENTV_PROVIDER_PROXY_TOKEN` - Bearer token for authentication
 
-The `createTargetClient()` function reads these automatically.
+The `createProviderClient()` function reads these automatically.
 
 ## Running
 

examples/features/script-grader-with-llm-calls/evals/contextual-precision.eval.yaml

@@ -5,7 +5,7 @@ assert:
       - bun
       - run
       - ../scripts/contextual-precision.ts
-    target:
+    provider:
       max_calls: 10
 providers:
   - llm

examples/features/script-grader-with-llm-calls/evals/contextual-recall.eval.yaml

@@ -5,7 +5,7 @@ assert:
       - bun
       - run
       - ../scripts/contextual-recall.ts
-    target:
+    provider:
       max_calls: 15
 providers:
   - llm

examples/features/script-grader-with-llm-calls/scripts/contextual-precision.ts

@@ -12,10 +12,10 @@
  * Retrieval context is extracted from expected_output.tool_calls output,
  * which represents the expected agent behavior (calling a retrieval tool).
  *
- * Requires `target: { max_calls: N }` in the grader YAML config,
+ * Requires `provider: { max_calls: N }` in the grader YAML config,
  * where N >= number of retrieval context nodes to evaluate.
  */
-import { createTargetClient, defineScriptGrader } from 'agentv';
+import { createProviderClient, defineScriptGrader } from 'agentv';
 import { extractRetrievalContext } from './utils.js';
 
 interface RelevanceResult {
@@ -63,22 +63,22 @@ export default defineScriptGrader(async (input) => {
     };
   }
 
-  const target = createTargetClient();
+  const provider = createProviderClient();
 
-  if (!target) {
+  if (!provider) {
     return {
       score: 0,
       assertions: [
         {
-          text: 'Target not available - ensure `target` block is configured in grader YAML',
+          text: 'Provider proxy not available - ensure `provider` block is configured in grader YAML',
           passed: false,
         },
       ],
     };
   }
 
   // Step 1: Use batch invocation to determine relevance of each node
-  // Demonstrates target override - uses gemini-llm regardless of default target
+  // Demonstrates provider override - uses gemini-llm regardless of default provider
   const requests = retrievalContext.map((node, index) => ({
     question: `Determine if this retrieved context node is relevant to answering the question.
 
@@ -95,10 +95,10 @@ Is this node relevant to answering the question? Respond with JSON only:
 }`,
     systemPrompt:
       'You are a precise relevance grader for RAG systems. Determine if a retrieved node contains information useful for answering the given question. Output valid JSON only.',
-    target: 'gemini-llm', // Override: use gemini-llm for relevance checks
+    provider: 'gemini-llm', // Override: use gemini-llm for relevance checks
   }));
 
-  const responses = await target.invokeBatch(requests);
+  const responses = await provider.invokeBatch(requests);
 
   // Step 2: Parse relevance scores for each node
   const relevanceScores: boolean[] = [];