chore: import upstream snapshot with attribution

examples/config-dynamic-var/README.md

@@ -0,0 +1,60 @@
+# config-dynamic-var (Dynamic Variable Generation)
+
+Generate variable values at runtime using JavaScript or Python.
+
+You can run this example with:
+
+```bash
+npx promptfoo@latest init --example config-dynamic-var
+cd config-dynamic-var
+```
+
+## Environment Variables
+
+This example requires:
+
+- `OPENAI_API_KEY` - Your OpenAI API key
+
+## How It Works
+
+Reference a script with `file://` in your vars. The script runs at evaluation time and can access other variables:
+
+```yaml
+tests:
+  - vars:
+      role: support
+      query: 'My order is late'
+      system_prompt: file://load_prompt.py # Loads support-specific prompt
+```
+
+The script receives the test context and returns the value:
+
+```python
+def get_var(var_name, prompt, other_vars):
+    role = other_vars.get("role")
+    return {"output": PROMPTS[role]}
+```
+
+## Function Signatures
+
+**Python** - Define `get_var(var_name, prompt, other_vars)`:
+
+```python
+def get_var(var_name, prompt, other_vars):
+    return {"output": "value"}  # or {"error": "message"}
+```
+
+**JavaScript** - Export a function:
+
+```javascript
+module.exports = async function (varName, prompt, otherVars, provider) {
+  return { output: 'value' }; // or { error: "message" }
+};
+```
+
+## Running the Example
+
+```bash
+promptfoo eval
+promptfoo view
+```

examples/config-dynamic-var/load_prompt.js

@@ -0,0 +1,20 @@
+/**
+ * Load system prompts dynamically based on agent role.
+ */
+
+const PROMPTS = {
+  support: `You are a customer support agent. Be empathetic, acknowledge
+the customer's frustration, and focus on resolving their issue quickly.`,
+
+  sales: `You are a sales representative. Be helpful and informative about
+our products and pricing. Don't be pushy - focus on understanding their needs.`,
+
+  technical: `You are a technical support engineer. Provide clear, accurate
+technical guidance. Include code examples or specific steps when relevant.`,
+};
+
+module.exports = async function (varName, prompt, otherVars) {
+  const role = otherVars.role || 'support';
+  const systemPrompt = PROMPTS[role] || PROMPTS.support;
+  return { output: systemPrompt };
+};

examples/config-dynamic-var/load_prompt.py

@@ -0,0 +1,17 @@
+"""Load system prompts dynamically based on agent role."""
+
+PROMPTS = {
+    "support": """You are a customer support agent. Be empathetic, acknowledge
+the customer's frustration, and focus on resolving their issue quickly.""",
+    "sales": """You are a sales representative. Be helpful and informative about
+our products and pricing. Don't be pushy - focus on understanding their needs.""",
+    "technical": """You are a technical support engineer. Provide clear, accurate
+technical guidance. Include code examples or specific steps when relevant.""",
+}
+
+
+def get_var(var_name, prompt, other_vars):
+    """Load the appropriate system prompt based on the role variable."""
+    role = other_vars.get("role", "support")
+    system_prompt = PROMPTS.get(role, PROMPTS["support"])
+    return {"output": system_prompt}

examples/config-dynamic-var/promptfooconfig.yaml

@@ -0,0 +1,36 @@
+# yaml-language-server: $schema=https://promptfoo.dev/config-schema.json
+description: 'Dynamic variables with JS/Python'
+
+prompts:
+  - |
+    {{system_prompt}}
+
+    User: {{query}}
+
+providers:
+  - openai:responses:gpt-5.4
+
+tests:
+  - vars:
+      role: support
+      query: 'Help! My mass spectrometer is making weird noises and sparking!'
+      system_prompt: file://load_prompt.py
+    assert:
+      - type: llm-rubric
+        value: empathetic and offers to help troubleshoot
+
+  - vars:
+      role: sales
+      query: 'How much for 10,000 rubber ducks? Asking for a friend.'
+      system_prompt: file://load_prompt.py
+    assert:
+      - type: llm-rubric
+        value: discusses pricing or bulk orders
+
+  - vars:
+      role: technical
+      query: 'My AI keeps responding in haiku. How do I make it stop?'
+      system_prompt: file://load_prompt.py
+    assert:
+      - type: llm-rubric
+        value: technical guidance with actionable steps