agentscope copilot-instructions.md

A code-review rule set for the AgentScope Python project. It covers code quality, security, testing, dependencies, and project conventions.

In plain words
What is it for?
Reviewing AgentScope changes, checking imports and internal names, spotting duplicated code, and verifying security and testing requirements.
Why use it?
It gives an agent clear standards for finding issues that could cause a pull request to be rejected or make the code harder to maintain.

Instructions file for GitHub Copilot

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 instructions/agentscope-ai/agentscope/copilot-instructions
Clone the repo
git clone --depth 1 https://github.com/agentscope-ai/agentscope

Made for: GitHub Copilot.

Per session 741 This file is loaded in full into every session.
When invoked 741 The same file — it is already loaded in full.
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 $0.00741 $0.00741
Opus 5 $0.00370 $0.00370
Sonnet 5 $0.00148 $0.00148
Haiku 4.5 $0.00074 $0.00074

Measured yesterday against content hash 85ea67ad5802, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agentscope copilot-instructions.md 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 yesterday.

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/copilot-instructions.md · 97 lines

How it starts

The opening of the file, as written. The whole thing — 97 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AgentScope Code Review Guide

You should conduct a strict code review. Each requirement is labeled with priority:

  • [MUST] must be satisfied or PR will be rejected
  • [SHOULD] strongly recommended
  • [MAY] optional suggestion

1. Code Quality

[MUST] Lazy Loading

  • Third-party library dependencies should be imported at the point of use, avoid centralized imports at file top
    • The Third-party library refers to libraries not included in the dependencies variable in pyproject.toml.
  • For base class imports, use factory pattern:
def get_xxx_cls() -> "MyClass":
    from xxx import BaseClass
    class MyClass(BaseClass): ...
    return MyClass

[SHOULD] Code Conciseness

After understanding the code intent, check if it can be optimized:

  • Avoid unnecessary temporary variables
  • Merge duplicate code blocks
  • Prioritize reusing existing utility functions

[MUST] Encapsulation Standards

  • All Python files under src/agentscope should be named with _ prefix, and exposure controlled through __init__.py
  • Classes and functions used internally by the framework that don't need to be exposed to users must be named with _ prefix

2. [MUST] Code Security

  • Prohibit hardcoding API keys/tokens/passwords
  • Use environment variables or configuration files for management
  • Check for debug information and temporary credentials
  • Check for injection attack risks (SQL/command/code injection, etc.)

3. [MUST] Testing & Dependencies

  • New features must include unit tests
  • New dependencies need to be added to the corresponding section in pyproject.toml
  • Dependencies for non-core scenarios should not be added to the minimal dependency list
  • Test assertions MUST compare the entire data structure rather than asserting individual fields, so reviewers can quickly understand the test intent. For random or non-deterministic fields, use AnyString and AnyValue from tests/utils.py for matching

4. Code Standards

[MUST] Comment Standards

  • Use English
  • All classes/methods must have complete docstrings, strictly following the template:
def func(a: str, b: int | None = None) -> str:
    """{description}

    Args:
        a (`str`):
            The argument a
        b (`int | None`, optional):
            The argument b

    Returns:
        `str`:
            The return str
    """
  • Use reStructuredText syntax for special content:
class MyClass:
    """xxx

    `Example link <https://xxx>`_

    .. note:: Example note

    .. tip:: Example tip

    .. important:: Example important info

    .. code-block:: python

        def hello_world():
            print("Hello world!")

    """

Read the full file on GitHub · 97 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. yesterday First seen · 97 lines · 741 tokens per session scan A 85ea67ad5802

Subscribe to this mod's changes

agentscope copilot-instructions.md is an instructions file published in the GitHub repository agentscope-ai/agentscope (30,227 stars, last pushed today), licensed Apache-2.0. It adds 741 tokens to every session, about $0.0037 per session on Opus 5. 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 2026-08-30.