maf-agent-skeleton

A standard project layout for building specialist agents under a Python source folder. It provides example files for prompts, tools, agent logic, a server, a container image, and tests.

In plain words
What is it for?
Use it when generating an agent with shared settings and GitHub Copilot integration, including its files, permissions handling, and test folder.
Why use it?
It gives developers a consistent starting structure when adding a new specialist agent to an existing codebase.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/microsoft/aks-lab-githubcopilot/maf-agent-skeleton
Any agent
npx skills add microsoft/AKS-Lab-GitHubCopilot --skill maf-agent-skeleton
Clone the repo
git clone --depth 1 https://github.com/microsoft/AKS-Lab-GitHubCopilot

Made for: Claude Code, Codex.

Per session 0 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,078 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.01078
Opus 5 $0.00000 $0.00539
Sonnet 5 $0.00000 $0.00216
Haiku 4.5 $0.00000 $0.00108

Measured 2d ago against content hash 8b3b9db672ae, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

maf-agent-skeleton scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.github/skills/maf-agent-skeleton/SKILL.md · 126 lines

How it starts

The opening of the file, as written. The whole thing — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Skill: MAF Agent Skeleton

Use this skill when generating any specialist agent under src/agents/<name>/.

Reference layout

src/agents/<name>/
├── __init__.py
├── prompts.py
├── tools.py
├── agent.py
├── server.py
├── Dockerfile
└── tests/__init__.py

agent.py template (specialist with MCP)

from __future__ import annotations

from dataclasses import dataclass

from agent_framework import AgentResponse
from agent_framework.github import GitHubCopilotAgent, GitHubCopilotOptions
from copilot.generated.rpc import PermissionDecisionApproveOnce
from copilot.session import PermissionRequestResult

from src.shared.copilot import build_copilot_client
from src.shared.settings import Settings
from .prompts import SYSTEM_PROMPT
from .tools import TOOLS  # usually empty: list[FunctionTool] = []


def _approve_all(_req: object, _ctx: dict[str, str]) -> PermissionRequestResult:
    return PermissionDecisionApproveOnce()


@dataclass(frozen=True)
class _RunnableAgent:
    """Adapter satisfying the shared `make_app` `_Runnable` protocol."""
    agent: GitHubCopilotAgent

    async def run(self, message: str, /) -> AgentResponse[None]:
        return await self.agent.run(message)


async def build_agent(settings: Settings) -> _RunnableAgent:
    agent = GitHubCopilotAgent(
        instructions=SYSTEM_PROMPT,
        client=build_copilot_client(),
        name="<name>",
        description="ZavaShop <name> specialist (...).",
        tools=list(TOOLS),
        default_options=GitHubCopilotOptions(
            model=settings.copilot_model,
            timeout=settings.copilot_timeout_seconds,
            on_permission_request=_approve_all,
            mcp_servers={
                "<key>": {
                    "type": "http",
                    "url": settings.<which>_mcp_url,
                    "tools": ["*"],
                    "timeout": int(settings.copilot_timeout_seconds * 1000),
                },
            },
        ),
    )
    return _RunnableAgent(agent=agent)

Read the full file on GitHub · 126 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 126 lines · 0 tokens per session scan A 8b3b9db672ae

Subscribe to this mod's changes

maf-agent-skeleton is a skill published in the GitHub repository microsoft/AKS-Lab-GitHubCopilot (7 stars, last pushed 28d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,078 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories