fix(template): update agent environment configuration and templates

hasansino · hasansino · commit 9e00e9cf99bd · 2025-09-24T17:17:40.000+03:00
diff --git a/pkg/agentenv/template/agentenv.yaml b/pkg/agentenv/template/agentenv.yaml
@@ -1,8 +1,6 @@
 ---
 
-# yaml-language-server: $schema=./agentenv.schema.json
-
-version: "1.0"
+version: 1.0
 
 project:
   name: go42x
@@ -48,8 +46,6 @@ providers:
       - workflows/40-debugging.tpl.md
       - workflows/50-investigation.tpl.md
       - workflows/60-code-review.tpl.md
-    agents:
-      - claude/agents/dummy.tpl.md
     tools:
       - Edit
       - Glob
@@ -201,7 +197,7 @@ mcp:
       - mcp__go42_kwb__list_files
 
   github:
-    enabled: true
+    enabled: false
     type: stdio
     name: github
     command: docker
@@ -303,7 +299,7 @@ mcp:
       - mcp__github__get_team_members
 
   gitlab:
-    enabled: true
+    enabled: false
     type: stdio
     name: gitlab
     command: docker
@@ -330,7 +326,7 @@ mcp:
       USE_PIPELINE: true
 
   jira:
-    enabled: true
+    enabled: false
     type: sse
     name: jira
     url: https://mcp.atlassian.com/v1/sse
@@ -350,7 +346,7 @@ mcp:
       - mcp__jira__assign_copilot_to_issue
 
   eleven-labs:
-    enabled: true
+    enabled: false
     type: stdio
     name: eleven-labs
     command: uvx
diff --git a/pkg/agentenv/template/chunks/20-operation.tpl.md b/pkg/agentenv/template/chunks/20-operation.tpl.md
@@ -1,6 +1,6 @@
 ### Operation
 
-- Ignore anything between `[IGNORE]` and `[/IGNORE]` tags - it used for internal notes
+- Ignore anything between `[IGNORE]` and `[/IGNORE]` tags in prompts.
 
 #### Execution Strategy
 
diff --git a/pkg/agentenv/template/chunks/60-responding.tpl.md b/pkg/agentenv/template/chunks/60-responding.tpl.md
@@ -1,5 +1,5 @@
 ### Responding
 
-The agent should respond in a concise and clear manner, directly addressing the user's query or request. It should avoid unnecessary elaboration or deviation from the topic at hand. The response should be structured logically, with a clear beginning, middle, and end, ensuring that the user can easily follow the information provided.
-
-When responding, the agent should also consider the user's context and prior interactions to provide relevant and personalized information. This includes referencing previous conversations or data points that may enhance the quality of the response.
+- The agent should respond in a concise and clear manner, directly addressing the user's query or request.
+- It should avoid unnecessary elaboration or deviation from the topic at hand.
+- The response should be structured logically, with a clear beginning, middle, and end, ensuring that the user can easily follow the information provided.
diff --git a/pkg/agentenv/template/chunks/70-tooling.tpl.md b/pkg/agentenv/template/chunks/70-tooling.tpl.md
@@ -2,6 +2,10 @@
 
 This is list of extra tools you are allowed to use to help you solve user requests.
 
+### Makefile
+
+Before using external tools, please check if the project has a Makefile and a relevant target for the task you are trying to accomplish. If it does, you should use make to run the target instead of using external tools.
+
 #### gh
 
 **git repo** - <https://github.com/cli/cli>
diff --git a/pkg/agentenv/template/chunks/80-search.tpl.md b/pkg/agentenv/template/chunks/80-search.tpl.md
@@ -114,20 +114,20 @@ What do you need to find?
 #### Search Tool Priority Order
 
 1. **Specialized MCP tools** (if available and applicable)
-   - Language servers for semantic understanding
-   - Knowledge base for indexed searches
-   - Domain-specific tools
+    - Language servers for semantic understanding
+    - Knowledge base for indexed searches
+    - Domain-specific tools
 
 2. **Native search tools** (always available)
-   - Grep for content search
-   - Glob for file patterns
-   - Read for direct access
-   - Task for complex searches
+    - Grep for content search
+    - Glob for file patterns
+    - Read for direct access
+    - Task for complex searches
 
 3. **Combination strategies**
-   - Use specialized tools for precision
-   - Use native tools for breadth
-   - Combine both for comprehensive analysis
+    - Use specialized tools for precision
+    - Use native tools for breadth
+    - Combine both for comprehensive analysis
 
 **Key principles:**
 - Start specific (specialized tools) then go broad (native tools)
diff --git a/pkg/agentenv/template/chunks/90-search-go.tpl.md b/pkg/agentenv/template/chunks/90-search-go.tpl.md
@@ -15,36 +15,36 @@ The Knowledge Base server excels at **fast text-based searches** and **initial c
 #### Best Use Cases
 
 1. **Initial Codebase Exploration**
-   - When you need to quickly understand what files exist in a project
-   - Getting an overview of the codebase structure
-   - Finding files by type (code, documentation, config)
-   ```
-   Example: "Show me all configuration files in the project"
-   Tool: mcp__go42x-kwb__list_files with type="config"
-   ```
+    - When you need to quickly understand what files exist in a project
+    - Getting an overview of the codebase structure
+    - Finding files by type (code, documentation, config)
+    ```
+    Example: "Show me all configuration files in the project"
+    Tool: mcp__go42x-kwb__list_files with type="config"
+    ```
 
 2. **Keyword and Pattern Search**
-   - Finding all occurrences of a specific string or pattern
-   - Searching for TODO comments, error messages, or specific text
-   - Looking for configuration values or environment variables
-   ```
-   Example: "Find all files mentioning 'database connection'"
-   Tool: mcp__go42x-kwb__search with query="database connection"
-   ```
+    - Finding all occurrences of a specific string or pattern
+    - Searching for TODO comments, error messages, or specific text
+    - Looking for configuration values or environment variables
+    ```
+    Example: "Find all files mentioning 'database connection'"
+    Tool: mcp__go42x-kwb__search with query="database connection"
+    ```
 
 3. **Quick File Content Retrieval**
-   - When you know the exact file path and need its contents
-   - Reading configuration files, documentation, or scripts
-   - Accessing non-Go files (YAML, JSON, Markdown, etc.)
-   ```
-   Example: "Show me the README file"
-   Tool: mcp__go42x-kwb__get_file with path="README.md"
-   ```
+    - When you know the exact file path and need its contents
+    - Reading configuration files, documentation, or scripts
+    - Accessing non-Go files (YAML, JSON, Markdown, etc.)
+    ```
+    Example: "Show me the README file"
+    Tool: mcp__go42x-kwb__get_file with path="README.md"
+    ```
 
 4. **Cross-Language Searches**
-   - Searching across mixed codebases (Go, JavaScript, Python, etc.)
-   - Finding patterns in build scripts, CI/CD configs, and documentation
-   - Exploring test fixtures and data files
+    - Searching across mixed codebases (Go, JavaScript, Python, etc.)
+    - Finding patterns in build scripts, CI/CD configs, and documentation
+    - Exploring test fixtures and data files
 
 #### Strengths
 
@@ -68,58 +68,58 @@ The Go Language Server provides **deep semantic analysis** and **code intelligen
 #### Best Use Cases
 
 1. **Understanding Go Code Structure**
-   - Analyzing package dependencies and imports
-   - Understanding module and workspace layout
-   - Getting package API summaries
-   ```
-   Example: "What packages does this project contain?"
-   Tool: mcp__gopls__go_workspace
-   ```
+    - Analyzing package dependencies and imports
+    - Understanding module and workspace layout
+    - Getting package API summaries
+    ```
+    Example: "What packages does this project contain?"
+    Tool: mcp__gopls__go_workspace
+    ```
 
 2. **Finding Symbol Definitions and References**
-   - Locating where a function, type, or variable is defined
-   - Finding all usages of a specific symbol
-   - Tracing method calls and type usage
-   ```
-   Example: "Find all references to the Server.Run method"
-   Tool: mcp__gopls__go_symbol_references with symbol="Server.Run"
-   ```
+    - Locating where a function, type, or variable is defined
+    - Finding all usages of a specific symbol
+    - Tracing method calls and type usage
+    ```
+    Example: "Find all references to the Server.Run method"
+    Tool: mcp__gopls__go_symbol_references with symbol="Server.Run"
+    ```
 
 3. **Semantic Code Search**
-   - Finding symbols by name with fuzzy matching
-   - Searching for types, interfaces, functions across packages
-   - Locating implementations of interfaces
-   ```
-   Example: "Find all types with 'Handler' in their name"
-   Tool: mcp__gopls__go_search with query="handler"
-   ```
+    - Finding symbols by name with fuzzy matching
+    - Searching for types, interfaces, functions across packages
+    - Locating implementations of interfaces
+    ```
+    Example: "Find all types with 'Handler' in their name"
+    Tool: mcp__gopls__go_search with query="handler"
+    ```
 
 4. **Code Context and Dependencies**
-   - Understanding file dependencies within a package
-   - Analyzing cross-file relationships
-   - Getting context about imports and usage
-   ```
-   Example: "What does server.go depend on?"
-   Tool: mcp__gopls__go_file_context with file="/path/to/server.go"
-   ```
+    - Understanding file dependencies within a package
+    - Analyzing cross-file relationships
+    - Getting context about imports and usage
+    ```
+    Example: "What does server.go depend on?"
+    Tool: mcp__gopls__go_file_context with file="/path/to/server.go"
+    ```
 
 5. **Package API Analysis**
-   - Understanding public APIs of packages
-   - Exploring third-party dependencies
-   - Reviewing exported types and functions
-   ```
-   Example: "Show me the public API of the storage package"
-   Tool: mcp__gopls__go_package_api with packagePaths=["example.com/storage"]
-   ```
+    - Understanding public APIs of packages
+    - Exploring third-party dependencies
+    - Reviewing exported types and functions
+    ```
+    Example: "Show me the public API of the storage package"
+    Tool: mcp__gopls__go_package_api with packagePaths=["example.com/storage"]
+    ```
 
 6. **Code Quality and Diagnostics**
-   - Finding compilation errors and issues
-   - Checking for type errors
-   - Validating code changes
-   ```
-   Example: "Check for errors in the edited files"
-   Tool: mcp__gopls__go_diagnostics with files=["/path/to/file.go"]
-   ```
+    - Finding compilation errors and issues
+    - Checking for type errors
+    - Validating code changes
+    ```
+    Example: "Check for errors in the edited files"
+    Tool: mcp__gopls__go_diagnostics with files=["/path/to/file.go"]
+    ```
 
 #### Strengths
 
diff --git a/pkg/agentenv/template/modes/50-voice.tpl.md b/pkg/agentenv/template/modes/50-voice.tpl.md
@@ -52,5 +52,5 @@ Use `eleven-labs` mcp server to generate voice responses.
 - Instead of: "Modified /src/components/Header.tsx line 42"
 - Say: "Updated the header component"
 
-- Instead of: "Error: TypeError at line 15"  
+- Instead of: "Error: TypeError at line 15"
 - Say: "Found a type error in the validation function"
diff --git a/pkg/agentenv/template/workflows/60-code-review.tpl.md b/pkg/agentenv/template/workflows/60-code-review.tpl.md
@@ -59,4 +59,5 @@ Comprehensive code evaluation for quality, security, and maintainability.
 - **🔴 Critical**: Must fix before merge
 - **🟡 Major**: Should fix before merge
 - **🔵 Minor**: Can fix in follow-up
-- **💡 Suggestion**: Consider for improvement
+- **💡 Suggestion**: Consider for improvement
+
