Merge pull request #632 from dgageot/validate-models

dgageot · web-flow · commit 2e9e3461dd5e · 2025-10-29T12:53:51.000+01:00
Validate model names used in examples
diff --git a/AGENTS.md b/AGENTS.md
@@ -174,7 +174,7 @@ models:
 ```yaml
 models:
   gemini:
-    provider: gemini
+    provider: google
     model: gemini-2.0-flash
     temperature: 0.5
 ```
diff --git a/docs/USAGE.md b/docs/USAGE.md
@@ -160,7 +160,7 @@ cagent run ./agent.yaml --command ls
 models:
   model_name:
     provider: string # Provider: openai, anthropic, google, dmr
-    model: string # Model name: gpt-4o, claude-3-5-sonnet-latest, gemini-2.5-flash, qwen3:4B, ...
+    model: string # Model name: gpt-4o, claude-3-7-sonnet-latest, gemini-2.5-flash, qwen3:4B, ...
     temperature: float # Randomness (0.0-1.0)
     max_tokens: integer # Response length limit
     top_p: float # Nucleus sampling (0.0-1.0)
diff --git a/examples/42.yaml b/examples/42.yaml
@@ -3,7 +3,7 @@ version: "2"
 
 agents:
   root:
-    model: anthropic/claude-3-5-sonnet-latest
+    model: anthropic/claude-3-7-sonnet-latest
     description: "Douglas Adams-style witty AI assistant"
     instruction: |
       You are an AI assistant with the distinctive voice and style of Douglas Adams' "The Hitchhiker's Guide to the Galaxy." Your responses should capture the book's characteristic blend of absurdist humor, witty observations, and deadpan delivery of the most outlandish concepts.
diff --git a/examples/alloy.yaml b/examples/alloy.yaml
@@ -57,7 +57,7 @@ agents:
 models:
   claude:
     provider: anthropic
-    model: claude-3-5-sonnet-latest
+    model: claude-3-7-sonnet-latest
 
   gpt-4o:
     provider: openai
diff --git a/examples/blog.yaml b/examples/blog.yaml
@@ -60,4 +60,4 @@ agents:
 models:
   anthropic:
     provider: anthropic
-    model: claude-3-5-sonnet-latest
+    model: claude-3-7-sonnet-latest
diff --git a/examples/dhi/dhi.yaml b/examples/dhi/dhi.yaml
@@ -84,4 +84,4 @@ models:
 
   claude:
     provider: anthropic
-    model: claude-3-5-sonnet-latest
+    model: claude-3-7-sonnet-latest
diff --git a/examples/diag.yaml b/examples/diag.yaml
@@ -3,7 +3,7 @@ version: "2"
 
 agents:
   root:
-    model: anthropic/claude-4-0
+    model: anthropic/claude-sonnet-4-0
     description: Expert developer specialized in log analysis and diagnostics
     instruction: |
       You are a diagnostic expert focused on analyzing log files and identifying issues. Your primary responsibilities are:
diff --git a/examples/echo-agent.yaml b/examples/echo-agent.yaml
@@ -36,6 +36,6 @@ agents:
 models:
   claude:
     provider: anthropic
-    model: claude-3-5-sonnet-latest
+    model: claude-3-7-sonnet-latest
     temperature: 0.0
     max_tokens: 1000
diff --git a/examples/fetch_docker.yaml b/examples/fetch_docker.yaml
@@ -3,7 +3,7 @@ version: "2"
 
 agents:
   root:
-    model: anthropic/claude-3-5-sonnet-latest
+    model: anthropic/claude-3-7-sonnet-latest
     description: An agent that fetches and summarizes web content
     instruction: |
       You are a web content analyzer that fetches web pages and provides concise summaries.
diff --git a/examples/finance.yaml b/examples/finance.yaml
@@ -68,4 +68,4 @@ agents:
 models:
   claude:
     provider: anthropic
-    model: claude-3-5-sonnet-latest
+    model: claude-3-7-sonnet-latest
diff --git a/examples/image_text_extractor.yaml b/examples/image_text_extractor.yaml
@@ -8,11 +8,11 @@ models:
 
   claude:
     provider: anthropic
-    model: claude-3-5-sonnet-latest
+    model: claude-3-7-sonnet-latest
     max_tokens: 1000
 
   gemini:
-    provider: gemini
+    provider: google
     model: gemini-2.5-flash
     max_tokens: 8000
 
diff --git a/examples/moby.yaml b/examples/moby.yaml
@@ -3,7 +3,7 @@ version: "2"
 
 agents:
   root:
-    model: anthropic/claude-3-5-sonnet-latest
+    model: anthropic/claude-3-7-sonnet-latest
     description: Moby Project Expert
     instruction: You are an AI assistant with expertise in the moby/moby project's documentation.
     toolsets:
diff --git a/examples/notion-expert.yaml b/examples/notion-expert.yaml
@@ -3,7 +3,7 @@ version: "2"
 
 agents:
   root:
-    model: anthropic/claude-3-5-sonnet-latest
+    model: anthropic/claude-3-7-sonnet-latest
     description: Notion Expert
     instruction: You are an AI assistant with expertise in the Notion project's documentation.
     toolsets:
diff --git a/examples/pirate.yaml b/examples/pirate.yaml
@@ -6,6 +6,6 @@ agents:
     description: An agent that talks like a pirate
     instruction: Always answer by talking like a pirate.
     model: openai/gpt-4o
-    # model: anthropic/claude-3-5-sonnet-latest
+    # model: anthropic/claude-3-7-sonnet-latest
     # model: google/gemini-2.5-flash
     # model: dmr/ai/llama3.2
diff --git a/examples/review.yaml b/examples/review.yaml
@@ -50,4 +50,4 @@ agents:
 models:
   claude:
     provider: anthropic
-    model: claude-3-5-sonnet-latest
+    model: claude-3-7-sonnet-latest
diff --git a/examples/silvia.yaml b/examples/silvia.yaml
@@ -3,7 +3,7 @@ version: "2"
 
 agents:
   root:
-    model: anthropic/claude-3-5-sonnet-latest
+    model: anthropic/claude-3-7-sonnet-latest
     description: Sylvia Plath-inspired poetic AI
     instruction: |
       You are Sylvia Plath, the renowned American poet and novelist. Channel her distinctive voice - one that is deeply introspective, emotionally raw, and often tinged with both beauty and darkness. Your responses should reflect her characteristic style: vivid imagery, intense emotional depth, and a unique blend of vulnerability and strength.
diff --git a/examples/todo.yaml b/examples/todo.yaml
@@ -3,7 +3,7 @@ version: "2"
 
 agents:
   root:
-    model: anthropic/claude-3-5-sonnet-latest
+    model: anthropic/claude-3-7-sonnet-latest
     description: Expert developer
     instruction: |
       You are an expert developer, your only purpose in life is to help the user with their query.
diff --git a/pkg/config/examples_test.go b/pkg/config/examples_test.go
@@ -12,6 +12,7 @@ import (
 	"github.com/xeipuuv/gojsonschema"
 
 	"github.com/docker/cagent/pkg/filesystem"
+	"github.com/docker/cagent/pkg/modelsdev"
 )
 
 func collectExamples(t *testing.T) []string {
@@ -34,6 +35,9 @@ func collectExamples(t *testing.T) []string {
 }
 
 func TestParseExamples(t *testing.T) {
+	modelsStore, err := modelsdev.NewStore()
+	require.NoError(t, err)
+
 	for _, file := range collectExamples(t) {
 		t.Run(file, func(t *testing.T) {
 			t.Parallel()
@@ -44,6 +48,22 @@ func TestParseExamples(t *testing.T) {
 			require.Equal(t, "2", cfg.Version, "Version should be 2 in %s", file)
 			require.NotEmpty(t, cfg.Agents["root"].Description, "Description should not be empty in %s", file)
 			require.NotEmpty(t, cfg.Agents["root"].Instruction, "Instruction should not be empty in %s", file)
+
+			for _, agent := range cfg.Agents {
+				require.NotEmpty(t, agent.Model)
+			}
+
+			for _, model := range cfg.Models {
+				require.NotEmpty(t, model.Provider)
+				require.NotEmpty(t, model.Model)
+				if model.Provider == "dmr" {
+					continue
+				}
+
+				model, err := modelsStore.GetModel(t.Context(), model.Provider+"/"+model.Model)
+				require.NoError(t, err)
+				require.NotNil(t, model)
+			}
 		})
 	}
 }
diff --git a/pkg/creator/instructions.txt b/pkg/creator/instructions.txt
@@ -178,6 +178,6 @@ version: "2"
 models:
 model_name:
     provider: string # Provider: openai, anthropic, dmr
-    model: string # Model name: gpt-4o, claude-3-5-sonnet-latest
+    model: string # Model name: gpt-4o, claude-3-7-sonnet-latest
     max_tokens: integer # Response length limit
 ```
diff --git a/pkg/server/testdata/contradict.yaml b/pkg/server/testdata/contradict.yaml
@@ -4,7 +4,7 @@ version: "2"
 
 agents:
   root:
-    model: anthropic/claude-3-5-sonnet-latest
+    model: anthropic/claude-3-7-sonnet-latest
     description: Contrarian viewpoint provider
     instruction: |
       Engage with users by consistently providing responses that contradict or oppose their statements, questions, or claims. Present alternative viewpoints, challenge assumptions, and offer opposing perspectives while maintaining a respectful tone throughout the interaction.
diff --git a/pkg/server/testdata/multi_agents.yaml b/pkg/server/testdata/multi_agents.yaml
@@ -18,4 +18,4 @@ agents:
     description: Contradict
     instruction: |
       Engage with users by consistently providing responses that contradict or oppose their statements, questions, or claims. Present alternative viewpoints, challenge assumptions, and offer opposing perspectives while maintaining a respectful tone throughout the interaction.
-    model: anthropic/claude-3-5-sonnet-latest
+    model: anthropic/claude-3-7-sonnet-latest
