agentmods.dev

OpenQuantum AGENTS.md

OpenQuantum AGENTS.md is an instructions file for Codex, OpenCode from xi-zhao/OpenQuantum. It costs 1,611 tokens per session, scanned A, original, MIT.

Architecture and development instructions for OpenQuantum, a research platform for AI agents working with quantum computing and scientific tasks. They explain how its runtime, skills, tools, connectors, and scientific checks are separated and combined.

In plain words
What is it for?
Use them when designing or implementing OpenQuantum features, especially agent workflows, tools, external tool connectors, scientific validation, or runtime behavior. They help decide where each piece of code belongs.
Why use it?
They help prevent different kinds of components from being mixed up or placed in the wrong part of the system. They also point developers to the project’s required architecture documents before making changes.

Instructions file for CodexOpenCode

48 5d ago A scan 1,611 tokens original MIT
View the source file xi-zhao/OpenQuantum
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/xi-zhao/openquantum/agents-md
Clone the repo
git clone --depth 1 https://github.com/xi-zhao/OpenQuantum

Made for: Codex, OpenCode.

Per session 1,611 This file is loaded in full into every session.
When invoked 1,611 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 $0.01611 $0.01611
Opus 5 $0.00805 $0.00805
Sonnet 5 $0.00322 $0.00322
Haiku 4.5 $0.00161 $0.00161

Measured 3d ago against content hash 82275f2a04e5, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

OpenQuantum 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 3d 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.

AGENTS.md · 93 lines

How it starts

The opening of the file, as written. The whole thing — 93 lines — stays where its author put it; the contents beside it link to each section on GitHub.

OpenQuantum

OpenQuantum 是一个科研 Agent 操作平台。DeepSeek Harness 负责通用 Agent Runtime，量子计算、 性质预测、案例咨询和通用问答等产品能力由 Agent Preset 组合独立的 Harness Skill 与 Tool Provider； Tool 可以由原生 Tool Plugin 注册，或由 MCP Server 暴露后经 Harness MCP Client 注册，必要时再调用 OpenQuantum Scientific Validator。只有 Tool hook 等 Agent 生命周期行为才在同一 Preset 组合受信任的 agent-scoped Host Plugin。

DeepSeek Harness 的“一切皆 Plugin”是运行时装配原则：可组合模块统一通过 Cordis Plugin row 注入、注册 Interface 并随 scope 回收。Skill、Tool、MCP Server、Validator 等是职责对象；它们说明模块负责什么， 不与 Cordis Plugin 这一装配机制竞争。

开始架构或实现工作前，先阅读 docs/README.md 的架构总览和权威导航；涉及架构基线、发布验收或当前风险时， 再阅读 docs/architecture/ARCHITECTURE_AUDIT.md。

四个职责面

下面四项用于判断规则归属，不表示一条 UI → Harness → 量子扩展 → Model 的线性调用链：

  1. UI：直接使用 Harness 原生 Web UI；浏览器展示通过 Harness Client Plugin、Slot 和 Harness RPC 扩展， 品牌壳由 deployment-scoped Host Plugin 使用 tapIndex 注入，OpenQuantum 特有设置经 Bounded Host Route 委托 Application Interface。
  2. Harness：Session、Turn、Step、Goal、Job、事件日志、工具调度、审批、沙箱和持久化。
  3. 量子扩展内容：Harness Skill 保存领域工作流和 Prompt；Tool 是 Agent 唯一调用的执行原语；Harness MCP Client 是 Tool Provider，MCP Server 通过协议暴露 Tool；Scientific Validator 产生科学 observations， Acceptance Profile 定义规则，只有 central Acceptance Builder 推导验收。
  4. Model：Provider route、模型能力元数据、协议适配、鉴权引用、超时和可用性探测。

不要新增独立的 OpenQuantum domain/platform Runtime。通用执行机制放 Harness；领域指令放原生 SKILL.md，确定性执行放 Tool implementation，可强制规则放 Validator 或可信 Host Plugin。相关源码可以为维护 locality 共置在 .agents/skills/<capability-id>/，但共置不表示 Harness 会自动绑定或启动这些模块。

扩展对象不可混用

  • Cordis Plugin：DSH 的统一装配与生命周期单元；按具体职责承载 Skill Provider、Tool Provider、 Harness MCP Client、Model Adapter、Host 或 Client 扩展，不能成为无边界的业务容器。
  • Skill：告诉 Agent 何时、为何和怎样做；不执行代码、不启动 MCP Server、不持有凭据。
  • Tool：Agent 可调用的原子动作；必须声明有界 schema、错误语义和副作用。
  • Tool Provider：在 Harness 侧把 Tool 注册进 Registry；典型实现是原生 Tool Plugin 或 Harness MCP Client。
  • MCP Server：通过协议暴露一个或多个 Tool；Agent 调用 MCP-exposed Tool，而不是“调用 MCP”。
  • Harness MCP Client：连接 MCP Server，并把其 Tool 注册进 Harness Tool Registry。
  • Bounded Host Route：只做请求边界与响应格式，然后委托 Application Interface；不复制业务规则。
  • Application Interface：统一拥有用例命令、状态转换、并发和安全规则，可被多个产品入口复用。
  • External API：Tool implementation 使用的下游网络合同；UI 和 Skill 不直接调用。
  • Scientific Validator：从结构化证据产生 observations；Acceptance Profile 定义规则，只有 central Acceptance Builder 推导最终 Acceptance。
  • Eval / Benchmark：开发、CI 和发布期证据，不进入用户请求的运行链。
  • Agent Preset：组合 Skill Provider、Tool Provider、Agent 策略，以及确有需要的 agent-scoped Host Plugin；不保存模型 Provider Route。
  • Deployment/Home Patch：组合 Provider Route、默认模型、默认 Agent Preset、deployment-scoped Host Plugin 与 Client Plugin。

Read the full file on GitHub · 93 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. 3d ago First seen · 93 lines · 1,611 tokens per session scan A 82275f2a04e5

Subscribe to this mod's changes

OpenQuantum AGENTS.md is an instructions file published in the GitHub repository xi-zhao/OpenQuantum (48 stars, last pushed 5d ago), licensed MIT. It adds 1,611 tokens to every session, about $0.0081 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.

Building agentsScientificagentic-aiai-agentdeepseekdeepseek-harnessmcp-servermodel-context-protocol
Related

Other instructions, from other repositories

sandbase-harness AGENTS.md

Instructions for sandbaseai/sandbase-harness, covering sandbase harness agent instructions, project identity, repository map, role a: project maintenance and issue workflow.

sandbaseai/sandbase-harness · 1,731 tokens

deepseek-harness-handbook AGENTS.md

Instructions for sandbaseai/deepseek-harness-handbook, covering contributor instructions, product boundary, content workflow, safety and project handoff and promotion playbook.

sandbaseai/deepseek-harness-handbook · 1,999 tokens

cdc-health-mcp-server CLAUDE.md

Instructions for cyanheads/cdc-health-mcp-server, covering agent protocol, domain, api surface, socrata api endpoints and cdc wonder api.

cyanheads/cdc-health-mcp-server · 8,242 tokens

dsh.fish CLAUDE.md

Instructions for stvlynn/dsh.fish, covering agentic coding guidelines, ⚠️ before you start, documentation map, understand the project and write frontend code.

stvlynn/dsh.fish · 2,064 tokens

dsh-oomol AGENTS.md

Instructions for oomol-lab/dsh-oomol, a project described as: OOMOL Connector for DeepSeek Harness—discover connected apps and execute Actions through progressive MCP discovery without exposing provider credentials.

oomol-lab/dsh-oomol · 153 tokens

aurora-agent AGENTS.md

AGENTS.md instructions for AURORA-NEURO/aurora-agent, covering aurora agent, what we can never compromise on, 1. honest labelling is the product, 2. invariants belong in the type system, not in comments and 3. determinism, byte for byte.

AURORA-NEURO/aurora-agent · 2,081 tokens