Feat: Code Undertaker Agent - New Agent Implementation

[DannylSyph3r]

User description

The Dead Code Undertaker

The Undertaker is a dead code detection agent that identifies unused functions, classes, variables, imports, methods and unreachable code across your entire repository using static analysis.

What It Does

• Scans all source files and detects unused code elements
• Provides deterministic confidence scoring (50-100%) for safe removal decisions
• Outputs clean JSON results for CI/CD integration

Use Case

Dead code accumulates naturally as features are refactored or deprecated. The Undertaker gives teams reliable, data-driven insights to make informed cleanup decisions.

Files Added

• agent.toml - Complete agent configuration
• agent.yaml - YAML format variant
• README.md - Full documentation with examples and best practices

Testing

Tested with a sample project to validate dead code detection accuracy across multiple file types and code patterns. Full testing walkthrough available in the demo video.

Documentation

Complete documentation included in README.md with usage examples, configuration options, confidence scoring rules, and integration guidance.

Breaking Changes

None. This is a new agent with no impact on existing functionality.

Demo

See the walkthrough: https://x.com/Slethware/status/1978785991314780402?t=hLnTNGKCapBhVRaxQqnaaQ&s=19

---

PR Type

Enhancement

---

Description

• Add new "Undertaker" agent for dead code detection with confidence scoring

• Implement comprehensive static analysis for unused functions, classes, variables, imports, and unreachable code

• Provide dual configuration formats (TOML and YAML) with identical functionality

• Include detailed documentation with usage examples, confidence scoring rules, and integration guidance

---

Diagram Walkthrough

flowchart LR
  config["Agent Configuration<br/>(TOML/YAML)"] --> analysis["Static Code Analysis"]
  analysis --> detection["Dead Code Detection"]
  detection --> scoring["Confidence Scoring<br/>(50-100%)"]
  scoring --> output["JSON Output<br/>(dead_code_analysis.json)"]

Loading

File Walkthrough

	Relevant files
Documentation	README.md Complete documentation for Undertaker agent                            agents/undertaker/README.md Add comprehensive documentation for the Undertaker dead code detection agent Document confidence scoring rules, analysis process, and output format Provide usage examples, integration guidance, and best practices Include technical details, limitations, and error handling information +189/-0
Configuration changes	agent.toml Agent configuration in TOML format                                              agents/undertaker/agent.toml Define agent configuration in TOML format with version 1.0 Implement detailed instructions for dead code detection using static analysis Configure arguments for min_confidence threshold and include_tests option Define comprehensive JSON output schema with summary statistics and dead code items Specify tools (git, filesystem, ripgrep) and execution strategy (plan) +164/-0  agent.yaml Agent configuration in YAML format                                              agents/undertaker/agent.yaml Define agent configuration in YAML format with identical functionality to TOML Implement same instructions, arguments, tools, and execution strategy Provide equivalent JSON output schema for dead code analysis results Mirror confidence scoring rules and analysis process from TOML version +178/-0

---

[qodo-merge-for-open-source]

PR Compliance Guide 🔍

Below is a summary of compliance checks for this PR:

Security Compliance
🟢	No security concerns identified No security vulnerabilities detected by AI analysis. Human verification advised for critical code.
Ticket Compliance
⚪	🎫 No ticket provided - [ ] Create ticket/issue <!-- /create_ticket --create_ticket=true --> </details></td></tr>
Codebase Duplication Compliance
⚪	Codebase context is not defined Follow the guide to enable codebase context checks.
Custom Compliance
⚪	No custom compliance provided Follow the guide to enable custom compliance check.

Compliance status legend

🟢 - Fully Compliant
🟡 - Partial Compliant
🔴 - Not Compliant
⚪ - Requires Further Human Verification
🏷️ - Compliance label

[qodo-merge-for-open-source]

PR Code Suggestions ✨

Explore these optional code suggestions:

Category	Suggestion	Impact
High-level	Use dedicated static analysis tools Replace the current unreliable static analysis method, which uses an LLM with ripgrep, with dedicated static analysis tools like ts-prune or vulture. These tools use Abstract Syntax Trees (ASTs) for more accurate and robust dead code detection. Examples: agents/undertaker/agent.toml [7-61] instructions = """ You are a dead code detection agent. Find unused code elements in the repository using static analysis. CORE MISSION: Identify unused functions, classes, variables, imports, types, enums, methods, and unreachable code. Provide confidence scoring based on reference counts and export status. Be reliable and dependable within reasonable limits. Output clean JSON results for integration. ANALYSIS PROCESS: ... (clipped 45 lines) agents/undertaker/agent.yaml [7-61] instructions: | You are a dead code detection agent. Find unused code elements in the repository using static analysis. CORE MISSION: Identify unused functions, classes, variables, imports, types, enums, methods, and unreachable code. Provide confidence scoring based on reference counts and export status. Be reliable and dependable within reasonable limits. Output clean JSON results for integration. ANALYSIS PROCESS: ... (clipped 45 lines) Solution Walkthrough: Before: # agent.toml (instructions for LLM) # 1. Discover all source files using filesystem tool. # 2. For each language, use ripgrep with regex patterns # to find all function/class definitions. # 3. For each definition found: # a. Use ripgrep again to count references in the codebase. # b. Use ripgrep to check for export keywords. # 4. Apply confidence rules based on reference count and export status. # 5. Format results into JSON. After: # agent.toml (instructions for LLM) # 1. Detect project languages (e.g., TypeScript, Python). # 2. Based on language, execute the appropriate static analysis tool: # - If TypeScript, run 'ts-prune' tool. # - If Python, run 'vulture' tool. # - ... other languages # 3. Capture the structured output from the tool. # 4. Transform the tool's output into the required JSON schema. # 5. Apply confidence scoring and filtering based on the parsed results. Suggestion importance[1-10]: 10 __ Why: This suggestion correctly identifies a fundamental architectural flaw in using ripgrep for static analysis and proposes a robust, industry-standard alternative that would dramatically improve the agent's reliability and correctness.	High
More