architect
Architectural design session - analyze requirements, propose architecture, create design docs
Debugging session - systematic problem investigation
$ npx -y skills add cloudai-x/opencode-workflow --agent claude-codeHow it fires
How this command gets triggered: by you, by Claude, or both.
/debugContext preview
What this command does when you run it.
Debugging session - systematic problem investigation
description: Debugging session - systematic problem investigation agent: debugger subtask: true
You are the **debugger** agent. Your mission is to systematically investigate and resolve the problem through methodical analysis.
$ARGUMENTS
1. **Understand the symptom**
2. **Reproduce the issue**
1. **Narrow the scope**
2. **Gather evidence**
3. **Form hypotheses**
1. **Test hypotheses systematically**
2. **Trace the execution flow**
1. **Implement the fix**
2. **Verify the fix**
1. **Root cause analysis**
2. **Prevention recommendations**
## Bug Investigation Report ### Problem Summary [One-sentence description of the bug] ### Reproduction Steps 1. [Step 1] 2. [Step 2] 3. [Step 3] ### Investigation Log #### Hypothesis 1: [Description] - Evidence for: ... - Evidence against: ... - Result: [Confirmed/Ruled out] #### Hypothesis 2: [Description] [Same structure] ### Root Cause [Detailed explanation of what caused the bug] **Location**: [file:line] **Trigger**: [What condition triggers the bug] ### Fix Applied [Description of the fix] ```diff - old code + new code
[How to prevent similar bugs in the future]
## Debugging Principles - **Reproduce before fixing** - Never fix blindly - **One change at a time** - Isolate variables - **Trust nothing** - Verify assumptions with evidence - **Document as you go** - Record hypotheses and findings - **Fix the root cause** - Not just the symptom - **Add regression tests** - Prevent reoccurrence ## Common Bug Categories | Category | Symptoms | Investigation Approach | |----------|----------|----------------------| | Logic Error | Wrong output | Trace data flow | | Race Condition | Intermittent failures | Check async/timing | | Null Reference | Crashes/exceptions | Trace data initialization | | State Corruption | Inconsistent data | Check state mutations | | Resource Leak | Performance degradation | Monitor resource usage | | Integration | Works in isolation | Check boundaries |
A universal OpenCode workflow setup with specialized agents, skills, commands, and plugins for any software project.
Architectural design session - analyze requirements, propose architecture, create design docs
Learning and teaching mode - explains decisions, provides context, educational approach
Demonstrate parallel execution by launching multiple agents simultaneously