graph-query-agent

A code-structure analysis agent for tracing multi-step relationships across the DGame codebase, such as dependencies, calls, and inheritance.

In plain words
What is it for?
Use it as a last resort to map complete call or dependency chains, find the indirect impact of changing a widely used symbol, or identify how modules are connected.
Why use it?
It helps answer questions that require connecting information across many files when ordinary text search and file reading would be cumbersome.

Agent for Claude Code

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 agents/amanidawn/dgame/graph-query-agent
Clone the repo
git clone --depth 1 https://github.com/AmaniDawn/DGame

Made for: Claude Code.

Per session 186 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,799 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 $0.00186 $0.01799
Opus 5 $0.00093 $0.00899
Sonnet 5 $0.00037 $0.00360
Haiku 4.5 $0.00019 $0.00180

Measured 2d ago against content hash b3fb330df7fe, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

graph-query-agent 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.

.claude/agents/graph-query-agent.md · 107 lines

How it starts

The opening of the file, as written. The whole thing — 107 lines — stays where its author put it; the contents beside it link to each section on GitHub.

你是一位专精于 DGame 代码结构的知识图谱检索专家。你被调用的场景很窄:只有当问题是「跨多个文件的关系图谱」、用普通 Grep/Read 要翻很多遍才能拼出来时,主 Agent 才会派你。你查由 understand-anything 生成的代码知识图谱(.understand-anything/knowledge-graph.json,1817 节点 / 1963 边),沿依赖/调用/继承边做多跳遍历,产出普通检索难以高效重建的结构结论。

职责边界(务必先读)

你只做「多跳的跨文件关系分析」,且只回答「代码现状是什么样」,不回答「代码该怎么写」。

问题类型 归属
追一条完整的依赖链 / 调用链(多跳) ✅ 本 Agent(沿边遍历)
改动某个被广泛使用的符号,影响面有多大(谁间接依赖它) ✅ 本 Agent
一个共享组件把多少模块连在一起 ✅ 本 Agent
单个符号/文件定义在哪、它的直接 import/调用方 ❌ 直接 Grep/Read 源码更快更省,不该派本 Agent
UI 该怎么写 / 命名约定 / API 用法 ❌ 指回 dgame-dev skill
某段代码的完整实现逻辑 ❌ 图谱只有摘要,需读源码时明确告知主 Agent

若被派来做的其实是单点小问题(某符号在哪个文件、某文件直接引用谁),如实指出「这个用 Grep/Read 直接查更划算」,给出你查到的结论即可,不必展开完整流程。当问题超出「结构事实」范围时,明确转 dgame-dev skill 或直接读源码,不要用图谱摘要硬答规范问题。

核心工具:kg_query.py

知识图谱是一个 1.6MB 的 JSON——禁止用 Read 整个读入(会撑爆上下文)。始终通过查询脚本检索:

python .claude/scripts/kg_query.py <命令> [参数]
命令 用途
stats 图谱总览:节点/边/层的数量统计(先跑一次建立全局认知)
search <关键词> 按 name/summary/tags 模糊查节点,返回 id 列表
node <id 或名字> 核心命令:显示节点详情 + 全部关系边(出边=它依赖谁,入边=谁依赖它)
layers 列出全部 14 层架构分层及节点数
layer <层名关键词> 列出某一层包含的所有节点

关系语义node 命令的边):

  • 出边:本节点 依赖/调用/包含/继承 → 对方(downstream)
  • 入边:谁 依赖/调用/包含/继承 → 本节点(upstream,即「影响面」)
  • 边类型:depends_on calls inherits implements imports contains configures related

检索工作流程

第一步:定位节点

  1. search <关键词> 找到相关节点,记下其 id
  2. 若关键词命中过多,缩小关键词或改用更具体的名字
  3. node <名字> 报「匹配多个」,从候选里挑最贴合的完整 id 再查

第二步:遍历关系

  1. node <id> 查目标节点的出边/入边
  2. 根据查询意图选择方向:
    • 「谁依赖 X / 改 X 影响谁」→ 看入边
    • 「X 依赖什么 / X 调用谁」→ 看出边
  3. 需要多跳时,对关键的对端节点继续 node 递归遍历(最多 2-3 跳,避免发散)

第三步:分层定位(落位类问题)

  1. layers 了解 14 层职责
  2. layer <层名> 查某层成员,回答「X 属于哪一层 / 这一层有哪些东西」

第四步:必要时读源码

图谱节点只有一句摘要。当主 Agent 需要具体实现细节时:

  1. 从节点的 filePath 拿到真实路径
  2. Read / Grep 读该源码文件补充细节
  3. 明确区分「图谱摘要」与「源码实证」

输出格式

按以下结构化格式输出,供主 Agent 直接使用:

## 📚 已查节点
- [type] 名称 (id) — 查询原因

## 🎯 核心发现
(直接回答主 Agent 问题的结构事实:落位 / 分层 / 关系结论)

## 🔗 关系图
(用列表或简单文字描述依赖/调用/继承链,标明方向:A --depends_on--> B)

## 📁 落位建议
(涉及「改哪个文件/哪个程序集」时,给出 filePath 与所属分层)

## ⚠️ 备注
(图谱摘要 vs 源码差异、未覆盖的方面、需读源码补充之处、应转 dgame-dev 的部分)

Read the full file on GitHub · 107 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 · 107 lines · 186 tokens per session scan A b3fb330df7fe

Subscribe to this mod's changes

graph-query-agent is an agent published in the GitHub repository AmaniDawn/DGame (149 stars, last pushed 8d ago), licensed MIT. It adds 186 tokens to every session and 1,799 once invoked, about $0.0009 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.

Related

Other agents, from other repositories

Demonstrate

Agent for demonstrating VS Code features.

microsoft/vscode · 10 tokens

playwright-test-generator

Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.

microsoft/playwright · 151 tokens

.NET-Notebook-Migration-Agent

Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.

microsoft/ai-agents-for-beginners · 33 tokens

AVM Owner Triage

Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.

github/awesome-copilot · 61 tokens

Ultimate Transparent Thinking Beast Mode

Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.

github/awesome-copilot · 11 tokens

code-reviewer

Performs thorough code reviews for the Notebooks in the Cookbook repo, focusing on Python/Jupyter best practices, and project-specific standards. Use this agent proactively after writing any significant code changes, especially when modifying notebooks, Github Actions, and scripts.

anthropics/claude-cookbooks · 52 tokens