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.
npx agentmods add instructions/agentinsight/agentinsight-sdk-python/agents-mdgit clone --depth 1 https://github.com/AgentInsight/agentinsight-sdk-pythonWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.06332 | $0.06332 |
| Opus 5 | $0.03166 | $0.03166 |
| Sonnet 5 | $0.01266 | $0.01266 |
| Haiku 4.5 | $0.00633 | $0.00633 |
Grade A, and why
agentinsight-sdk-python AGENTS.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.
How it starts
The opening of the file, as written. The whole thing — 442 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AgentInsight Python SDK — 智能体执行规则
本文档是AI智能体（AI Agent）执行本项目开发任务的规则依据。智能体在执行任何任务前，必须先完整阅读本文档，并严格按照本文档定义的规则执行。
1. 项目基本信息
1.1 技术栈
| 类别 | 技术选型 |
|---|---|
| 语言/框架 | Python >= 3.10, < 4.0 |
| 包管理 | uv 0.11.2（锁定依赖 + 构建后端） |
| 类型检查 | mypy >= 1.0（严格模式，disallow_untyped_defs = true） |
| 代码风格 | ruff >= 0.15.2（Google docstring 约定） |
| HTTP 客户端 | httpx >= 0.15.4 |
| 数据模型 | Pydantic >= 2 |
| 可观测性 | OpenTelemetry API + SDK >= 1.33.1（TracerProvider + OTLP HTTP Exporter） |
| API 客户端 | Fern 自动生成（agentinsight/api/，不可手改） |
| 测试 | pytest >= 7.4 + pytest-xdist + pytest-asyncio + pytest-httpserver |
| 重试/退避 | backoff + tenacity |
| 猴子补丁 | wrapt（OpenAI 集成） |
| 构建 | uv build --no-sources |
| 文档 | pdoc >= 15.0.4（Google 格式） |
| 部署 | PyPI 发布（GitHub Actions 自动触发） |
1.2 项目关键特征
- OpenTelemetry 原生：SDK 完全构建在 OpenTelemetry 标准之上，不作私有 tracing 协议。Span 通过
OTLPSpanExporter发送到 AgentInsight OTLP 端点。 - 单例资源管理：
AgentInsightResourceManager按public_key管理单例，每个 API key 维护独立的TracerProvider、AgentInsightSpanProcessor和后台线程。 - 多项目安全隔离：通过
ContextVar实现多项目场景下的客户端隔离，防止 trace 数据跨项目泄漏。 - 三级集成方式：直接 SDK（
AgentInsight客户端）、@observe装饰器、OpenAI 直插替换（agentinsight.openai）、LangChain 回调处理器。 - 后台任务线程：
MediaUploadConsumer（媒体上传）+ScoreIngestionConsumer（评分摄入），通过 sentinel 对象实现优雅关闭。 - 自动生成 API 客户端：
agentinsight/api/由 Fern 根据 OpenAPI 规范自动生成，覆盖 ingestion、prompts、datasets、scores、media 等全部端点。
1.3 项目目录结构
| 目录 | 说明 |
|---|---|
agentinsight/__init__.py |
公共 API 入口，__all__ 导出全部公开符号 |
agentinsight/_client/ |
核心 SDK：客户端、装饰器、资源管理器、Span 处理器、Span 过滤器、属性传播、数据集 |
agentinsight/_task_manager/ |
后台消费者：媒体上传队列与管理、评分摄入 |
agentinsight/_utils/ |
内部工具：HTTP 客户端、序列化器、提示缓存、错误日志、环境变量 |
agentinsight/api/ |
Fern 自动生成的 API 客户端，禁止手改 |
agentinsight/openai.py |
OpenAI SDK 直插式集成（wrapt 猴子补丁） |
agentinsight/langchain/ |
LangChain 回调处理器集成 |
tests/unit/ |
单元测试：不需要 AgentInsight 服务器，使用 InMemorySpanExporter |
tests/e2e/ |
端到端测试：需要真实 AgentInsight 服务器，使用带轮询重试的 API wrapper |
tests/live_provider/ |
供应商测试：需要真实 OpenAI/LangChain API 调用 |
tests/support/ |
e2e 共享辅助工具（API wrapper、重试、轮询） |
scripts/select_e2e_shard.py |
CI 分片选择器：历史权重 + 贪心装箱算法 |
scripts/codex/ |
Codex 云/工作树引导和共享快速检查 |
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.
- yesterday First seen · 442 lines · 6,332 tokens per session scan A f5de99795651
agentinsight-sdk-python AGENTS.md is an instructions file published in the GitHub repository AgentInsight/agentinsight-sdk-python (10 stars, last pushed 1mo ago), licensed MIT. It adds 6,332 tokens to every session, about $0.0317 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 2026-08-31.
Other instructions, from other repositories
mini-swe-agent copilot-instructions.md
Instructions for SWE-agent/mini-swe-agent, covering style guide, test style, bad and good.
openagent CLAUDE.md
Claude Code instructions for the-open-agent/openagent, covering claude.md, commands, architecture, backend (go / beego) and frontend (react).
Octop AGENTS.md
Instructions for TencentCloud/Octop, covering agents.md, 1. collaboration principles, think before writing, simplicity first and surgical edits.
intelligent-terminal rust.instructions.md
Concise Rust coding conventions for this repository.
neeko AGENTS.md
Instructions for tincopper/neeko, covering neeko — repository guidelines, 项目概览, 目录结构, 前端（feature-based 架构） and 后端（domain-driven 模块化架构）.
lime AGENTS.md
AGENTS.md instructions for limecloud/lime, covering lime 源码仓库指南, 事实源, 工程约束, 协议与质量 and 执行方式.