Add collect tool type

README.md

@@ -69,9 +69,9 @@ sources:
       overlap_lines: 10
 ```
 
-### Tools
+### Search Tools
 
-Each tool maps to a source and defines the MCP tool interface:
+Each search tool maps to a source and defines the MCP tool interface:
 
 ```yaml
 tools:
@@ -83,6 +83,36 @@ tools:
     result_format: docs
 ```
 
+### Collect Tools
+
+Collect tools let agents write structured data back to the server. Unlike search tools, they don't query anything — they validate the agent's input against a YAML-defined schema and store it as JSONB in the database. Use them to gather signal from agents without writing any code.
+
+The first built-in use case is search feedback: agents report whether search results were helpful, what they tried, and what went wrong. This surfaces broken or misleading documentation quickly. But collect tools are generic — you can define any schema for any use case (e.g., broken link reporting, feature requests, error logging).
+
+```yaml
+tools:
+  - name: submit-feedback
+    type: collect
+    description: "Submit feedback on whether search results were helpful."
+    response: "Feedback recorded. Thank you."
+    schema:
+      tool_name:
+        type: string
+        description: "Which search tool was used"
+        required: true
+      rating:
+        type: enum
+        values: ["helpful", "not_helpful"]
+        description: "Whether the results were helpful"
+        required: true
+      comment:
+        type: string
+        description: "What worked or didn't work"
+        required: true
+```
+
+Each field in `schema` supports `type` (`string`, `number`, or `enum`), an optional `description` (shown to the agent), `required` (defaults to false), and `values` (required for `enum` fields). The validated input is written as JSONB to the `collected_data` table along with the tool name and a timestamp.
+
 ### Built-in Chunker Types
 
 | Type | Best For | Splits On |
@@ -187,6 +217,9 @@ docker compose up
 # Seed index
 docker compose exec app npx tsx scripts/seed-index.ts
 
+# Run unit tests
+npm test
+
 # Test search
 docker compose exec app npx tsx scripts/test-search.ts "your query"
 

mcp-docs.example.yaml

@@ -24,6 +24,27 @@ tools:
     max_limit: 20
     result_format: docs
 
+  # Collect tools write structured data from agents to the database.
+  # Define input schema, description, and a canned response — no code needed.
+  - name: submit-feedback
+    type: collect
+    description: "Submit feedback on whether search results were helpful."
+    response: "Feedback recorded. Thank you."
+    schema:
+      tool_name:
+        type: string
+        description: "Which search tool was used"
+        required: true
+      rating:
+        type: enum
+        values: ["helpful", "not_helpful"]
+        description: "Whether the results were helpful"
+        required: true
+      comment:
+        type: string
+        description: "What worked or didn't work"
+        required: true
+
 embedding:
   provider: openai
   model: text-embedding-3-small

mcp-docs.yaml

@@ -118,6 +118,29 @@ tools:
     max_limit: 20
     result_format: code
 
+  - name: submit-feedback
+    type: collect
+    description: "After using search results, you can submit feedback on whether the advice worked or not. Report what you tried and whether it succeeded or failed. If something failed but you later found the missing information elsewhere, mention that too. This feedback helps improve search quality."
+    response: "Feedback recorded. Thank you."
+    schema:
+      tool_name:
+        type: string
+        description: "Which search tool was used"
+        required: true
+      query:
+        type: string
+        description: "The original search query"
+        required: true
+      rating:
+        type: enum
+        values: ["helpful", "not_helpful"]
+        description: "Whether the results were helpful"
+        required: true
+      comment:
+        type: string
+        description: "What was tried, what failed/worked, what info was missing"
+        required: true
+
 embedding:
   provider: openai
   model: text-embedding-3-small