菜单
Improve GitHub Copilot custom agents by validating agent contracts, tightening tool and MCP skill routing, and minimizing prompt bloat while keeping handoffs bounded to real workspace agents. Use this whenever the user wants to create, debug, or refine a custom agent.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Improve a markdown prompt file using Agent Lightning APO (Automatic Prompt Optimization). Use when the user asks to optimize or improve a markdown prompt, or starts a message with /trainer-optimize.
Own the end-to-end trainer loop for agent contract targets (*.agent.md files, custom agent definitions, and agent instruction documents). Use this whenever the caller needs to research, synthesize datasets, optimize, validate, and write back a trained candidate for an agent-type target. Prefer this specialized loop whenever the selected target defines tool routing, MCP skill configuration, agent personas, or handoff behavior rather than raw prompts, code, or skill definitions.
Own the end-to-end trainer loop for Python code targets optimized with Microsoft Trace (nodes, bundles, models, and trainable agent components). Use this whenever the caller needs to research, synthesize test-based datasets, optimize, validate, and write back a trained candidate for a code-type target. Prefer this specialized loop for any Python file or callable that benefits from deterministic, test-based or benchmark-based feedback rather than open-ended language instruction quality.
Own the end-to-end trainer loop for Python code targets optimized with Microsoft Trace (nodes, bundles, models, and trainable agent components). Use this whenever the caller needs to research, synthesize test-based datasets, optimize, validate, and write back a trained candidate for a code-type target. Prefer this specialized loop for any Python file or callable that benefits from deterministic, test-based or benchmark-based feedback rather than open-ended language instruction quality.
Own the end-to-end trainer loop for prompt-like files (*.prompt.md, *.prompty, *.instructions.md, system prompts, and other natural-language instruction artifacts). Use this whenever the caller needs to research, synthesize datasets, optimize, validate, and write back a trained candidate for a prompt-type target. Prefer this specialized loop for any file whose primary content is natural-language instructions rather than code, skill configuration, or agent contracts.
Own the end-to-end trainer loop for prompt-like files (*.prompt.md, *.prompty, *.instructions.md, system prompts, and other natural-language instruction artifacts). Use this whenever the caller needs to research, synthesize datasets, optimize, validate, and write back a trained candidate for a prompt-type target. Prefer this specialized loop for any file whose primary content is natural-language instructions rather than code, skill configuration, or agent contracts.
| name | engineer-copilot-agent |
| description | Improve GitHub Copilot custom agents by validating agent contracts, tightening tool and MCP skill routing, and minimizing prompt bloat while keeping handoffs bounded to real workspace agents. Use this whenever the user wants to create, debug, or refine a custom agent. |
| argument-hint | Describe the target custom agent, whether the concern is triggering, routing, handoffs, structure, evals, or all, and any observed failures such as stale tool names, bad handoffs, or bloated instructions. |
| license | MIT |
| compatibility | Python 3.11+. Works in repositories that store GitHub Copilot custom agents as `.agent.md` files alongside reusable skills. |
| metadata | {"author":"Tyler Kendrick","version":"0.1.0"} |
Use this skill to create or improve GitHub Copilot custom agents. Treat triggering, routing, handoffs, and prompt size as separate concerns. Keep the top-level agent contract lean.
Read references/copilot-agent-standard.md for the baseline contract and references/context-minimization-loop.md before editing.
Use it when the user wants to:
.agent.md fileDo not use this skill for prompt-only or skill-only work unless the target artifact is a Copilot custom agent contract or the user explicitly wants to design one.
.agent.md path, or a description of the agent to createFollow this order:
python scripts/discover_runtime_surface.py --repo-root <repo-root> --json.python scripts/validate_agent.py <agent-path> --repo-root <repo-root> --json.python scripts/analyze_agent_body.py <agent-path> --repo-root <repo-root> --json.references/ and mechanical checks to scripts/.Optimize discoverability and fit:
name, description, and hints specificName only tools, skills, helpers, and agents that actually exist in the workspace or live session.
Make ownership explicit:
Keep the .agent.md file as a routing layer.
references/scripts/Use a recursive pass until the prompt is only as large as needed:
Stop when each remaining section is necessary for triggering, routing, handoffs, or bounded execution.
When the user asks for evals, prefer coverage that catches:
When improving a custom agent, return: