endorphin-ai/hasbrains-agent-kit / agent-engineering/skills/convert-to-copilot
convert-to-copilot
VerifiedThis skill should be used when converting a Claude Code agent (+ its launch command + its skills) into a single GitHub Copilot agent (.github/agents/*.agent.md) with supporting context files under .copilot_utils/{agent_name}/. Load when someone says 'convert to copilot', 'make copilot agent', 'copilot version of this agent', 'port this agent to copilot', or when /convert-to-copilot is invoked with an agent file path.
| File | Kind | Size |
|---|---|---|
| README.md | markdown | 1.9 KB |
| SKILL.md | skill md | 11.3 KB |
| assets/copilot-agent-template.md | markdown | 1.6 KB |
| examples/pr-reviewer.md | markdown | 5.2 KB |
| references/advanced-patterns.md | markdown | 3.6 KB |
| references/advanced-subagents.md | markdown | 4.0 KB |
| references/conversion-body.md | markdown | 4.4 KB |
| references/conversion-output.md | markdown | 3.1 KB |
| references/conversion-phases.md | markdown | 4.6 KB |
| references/tool-mapping.md | markdown | 6.1 KB |
| scripts/claude-to-copilot.sh | script | 4.6 KB |
| scripts/convert-agents.sh | script | 1.3 KB |
| scripts/copilot-to-claude.sh | script | 5.6 KB |
| scripts/lib-extract.sh | script | 1.8 KB |
references/conversion-body.md
View on GitHub# Conversion Phase 4: Convert Agent Body
See also: `conversion-phases.md` (phases 1–3) · `conversion-output.md` (phases 5–7)
Convert the agent body section by section. Write result to `{output_folder}/4-body.md`.
Preserve all logic — only update references and tracking.
---
## 4.1 Role / Identity section
Keep as-is. Copy directly.
---
## 4.2 Skill usage table
Update from skill names to `.copilot_utils/{agent_name}/` paths:
**Before (Claude):**
```markdown
| Skill | What it knows | When you use it |
| ------------------ | -------------------- | --------------- |
| `review-triage` | Risk classification | Phase 2 |
```
**After (Copilot):**
```markdown
| Context file | What it knows | When you use it |
| ---------------------------------------------------- | ------------------- | --------------- |
| `.copilot_utils/{agent_name}/trees/review-triage.md` | Risk classification | Phase 2 |
```
---
## 4.3 Task tracking (MANDATORY conversion)
Replace every `TaskCreate`/`TaskUpdate` block with `manage_todo_list`.
**⚠️ CRITICAL: A bare JSON array `[{...}]` is NOT correct and will fail. Always use `"operation": "write"`.**
**Before:**
```python
task1_id = TaskCreate(subject="Phase 1: Identify input", activeForm="Identifying")
TaskUpdate(taskId=task1_id, status="in_progress")
# ... work ...
TaskUpdate(taskId=task1_id, status="completed")
```
**After — initial setup (one call before any work starts):**
```json
{
"operation": "write",
"todoList": [
{ "id": 1, "title": "Phase 1: Identify input and locate test", "status": "not-started" },
{ "id": 2, "title": "Phase 2: Create branch", "status": "not-started" }
]
}
```
**After — mark in-progress (before each phase):**
```json
{ "operation": "update", "updates": [{ "id": 1, "status": "in-progress" }] }
```
**After — mark completed (immediately after finishing):**
```json
{ "operation": "update", "updates": [{ "id": 1, "status": "completed" }] }
```
**Rules:**
- Always call `operation: write` with ALL todos BEFORE starting any work
- Update ONE task at a time — in-progress when starting, completed when done
- Sequential `addBlockedBy` chains → simple ascending ID order (1 before 2 before 3)
**Embed this PROGRESS TRACKING section in the agent body:**
````markdown
## PROGRESS TRACKING
**CRITICAL:** Use `manage_todo_list` tool to track progress.
**Initial Setup (before starting work):**
```json
{
"operation": "write",
"todoList": [
{ "id": 1, "title": "Phase 1: ...", "status": "not-started" },
{ "id": 2, "title": "Phase 2: ...", "status": "not-started" }
]
}
```
**During Execution:** Mark in-progress when starting, completed when done.
```json
{ "operation": "update", "updates": [{ "id": 1, "status": "in-progress" }] }
{ "operation": "update", "updates": [{ "id": 1, "status": "completed" }] }
```
Update ONE task at a time. Never batch multiple status changes.
````
---
## 4.4 Skill references in phase bodies
Replace Claude skill references with file paths:
| Before (Claude) | After (Copilot) |
| -------------------------------------------------- | ----------------------------------------------------------------------------- |
| `**Skill used:** \`review-triage\`` | `**Apply tree:** \`.copilot_utils/{agent_name}/trees/review-triage.md\`` |
| `"Follow the triage from \`review-triage\`"` | `"Read and follow \`.copilot_utils/{agent_name}/trees/review-triage.md\`"` |
| `"Using knowledge from the \`code-conventions\` skill"` | `"Read \`.copilot_utils/{agent_name}/context/code-conventions.md\`"` |
---
## 4.5 Script references
```
bash tests/.claude/skills/{name}/scripts/{script}.sh
→ bash .copilot_utils/{agent_name}/scripts/{script}.sh (if script was copied)
→ or inline the bash command directly
Skill("skill-name") calls → remove; replace with "Read `.copilot_utils/{agent_name}/context/{skill}.md`"
```
---
## 4.6 Memory section
```
Before: .claude/agent-memory/{agent-name}/MEMORY.md
After: .copilot_utils/{agent-name}/memory/MEMORY.md
```
---
## 4.7 Keep as-is
- Bash commands and git workflow
- MCP tool calls (already use full IDs)
- ALWAYS_DO / NEVER_DO rules
- Error handling blocks
- Return format / final report
- Phase numbering and structure
- Variable definitions and state tracking
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |