What is j6-plugin-graph-diff?

A tool for comparing two PyTorch FX Graph files, which describe the operations in a machine-learning model, and linking differences back to model source code.

Use it to compare training, evaluation, export, inference, or different-version graphs, then locate the related code differences and suggest fixes.

It avoids manual comparison of complex model graphs and helps distinguish changed operations, parameters, and structure.

j6-plugin-graph-diff — Skill by HorizonRobotics/OE-Skills

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.

Any agent

npx skills add HorizonRobotics/OE-Skills --skill j6-plugin-graph-diff

Clone the repo

git clone --depth 1 https://github.com/HorizonRobotics/OE-Skills

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/horizonrobotics/oe-skills/j6-plugin-graph-diff/github.svg)](https://agentmods.dev/skills/horizonrobotics/oe-skills/j6-plugin-graph-diff)

Your own site

<a href="https://agentmods.dev/skills/horizonrobotics/oe-skills/j6-plugin-graph-diff"><img src="https://agentmods.dev/badge/skills/horizonrobotics/oe-skills/j6-plugin-graph-diff/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/horizonrobotics/oe-skills/j6-plugin-graph-diff"><img src="https://agentmods.dev/badge/skills/horizonrobotics/oe-skills/j6-plugin-graph-diff.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 50 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 4,260 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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.

Model	Per session	Once invoked
Fable 5.1	$0.00050	$0.04260
Opus 5.5	$0.00020	$0.01704
Sonnet 5.5	$0.00010	$0.00852
Haiku 4.5	$0.00005	$0.00426

Measured 2mo ago against content hash cccef061e39d, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

j6-plugin-graph-diff 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 2mo ago.

The scan reads SKILL.md. This mod also ships 1 executable file (j6_plugin_graph_diff.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

horizon/skills/plugin/j6-plugin-graph-diff/SKILL.md · 441 lines

How it starts

The opening of the file, as written. The whole thing — 441 lines — stays where its author put it; the contents beside it link to each section on GitHub.

FX Graph 计算图对比与源码定位工具

强约束（必须遵守）

1) 必须调用外挂脚本，禁止 Agent 自行实现差异对比

本 Skill 的计算图差异对比必须通过调用外挂脚本 j6_plugin_graph_diff.py 完成，严禁 Agent 自己编写 Python 代码实现 diff 逻辑。

✅ 必须：通过 Bash 工具执行 python j6-plugin-graph-diff/j6_plugin_graph_diff.py --file1 ... --file2 ...
❌ 禁止：Agent 自行编写 difflib/正则匹配等代码来解析计算图
❌ 禁止：用 Bash 的 diff 命令替代外挂脚本
❌ 禁止：用 Read 工具读取两个文件后自行比较

原因：外挂脚本包含完整的 FX Graph 解析逻辑（节点解析、差异分类、相似度计算等），Agent 自行实现容易遗漏格式兼容性、分类精度等问题。

2) 外挂脚本调用前必须确认文件存在

在调用脚本前，Agent 必须确认两个输入文件路径存在且可读，避免脚本执行失败。

3) 脚本路径

外挂脚本位于：j6-plugin-graph-diff/j6_plugin_graph_diff.py

调用时使用相对路径：

python j6-plugin-graph-diff/j6_plugin_graph_diff.py --file1 <path1> --file2 <path2> -o <output_path>

目标

对比两份 FX Graph 计算图文件（例如评测与导出、训练与推理、不同版本之间的计算图），实现：

精确识别计算图差异：找出第一个不一致点、所有差异块（由外挂脚本完成）
差异分类统计：按差异类型（算子变化、参数变化、结构变化）进行分类
源码反向定位：根据 diff 报告，由 Agent 分析并将差异映射回原始模型代码位置

职责分工

本 Skill 采用外挂脚本 + Agent 分析的分工模式：

组件	职责	输入	输出
外挂脚本 `j6_plugin_graph_diff.py`	计算图差异对比	两个 FX Graph 文件	diff 报告（文本格式）
Agent Skill	源码定位与分析	diff 报告 + 模型源码	源码位置、修复建议

重要：Agent 不得自行实现外挂脚本的职责。差异对比必须调用外挂脚本，Agent 只负责解读脚本输出并进行源码定位。

外挂脚本职责（j6_plugin_graph_diff.py）

外挂脚本只负责计算图层面的差异对比，不涉及源码分析：

✅ 解析两个 FX Graph 文件
✅ 找出第一个差异点
✅ 找出所有差异块
✅ 对差异进行分类（operator_change / parameter_change / structure_change）
✅ 计算相似度
✅ 生成结构化 diff 报告
❌ 不负责源码反向定位

Agent Skill 职责

Agent 负责根据 diff 报告进行源码定位和问题分析：

✅ 调用外挂脚本生成 diff 报告
✅ 分析 diff 报告，理解差异内容
✅ 根据差异节点信息，在模型源码中搜索对应位置
✅ 提供源码上下文和修复建议
✅ 判断差异是否预期（结合用户提供的场景说明）

适用范围

本 Skill 适用于以下场景：

评测与导出对比：对比模型评测和导出阶段的计算图差异
训练与推理对比：对比训练模式和推理模式的计算图差异
版本对比：对比不同版本模型的结构变化
调试定位：快速定位计算图异常点，并映射回源码

核心能力

1. 计算图差异识别（外挂脚本）

✅ 识别算子类型变化（如 torch.relu → torch.leaky_relu）
✅ 识别算子参数变化（如 args=(x, 1) → args=(x, 0.5)）
✅ 识别结构变化（如新增/删除节点、改变连接关系）

2. 源码反向定位（Agent Skill）

✅ 解析 diff 报告，提取关键差异节点
✅ 在模型源码中搜索对应的操作（如模块名、函数名）
✅ 匹配 forward 函数中的调用位置
✅ 提供代码上下文（前后 N 行）便于理解
✅ 判断差异是否符合预期场景

Read the full file on GitHub · 441 lines

Files

What ships with it

2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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.

2mo ago First seen · 441 lines · 50 tokens per session scan A cccef061e39d

Subscribe to this mod's changes

j6-plugin-graph-diff is a skill published in the GitHub repository HorizonRobotics/OE-Skills (27 stars, last pushed 4mo ago), licensed Apache-2.0. It adds 50 tokens to every session and 4,260 once invoked, about $0.0002 per session on Opus 5.5. 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.

Data and AI Debugging

j6-plugin-graph-diff next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
j6-plugin-graph-diffHorizonRobotics/OE-Skills	Skill	27	4mo ago	Apache-2.0	A	50	Claude Code, Codex
debug-inferenceNVIDIA/OpenShell	Skill	21k	yesterday	Apache-2.0	A	120	Claude Code, Codex
notebooklmteng-lin/notebooklm-py	Skill	20k	yesterday	MIT	A	79	Claude Code, Codex
oh-my-poshJanDeDobbeleer/oh-my-posh	Skill	24k	2d ago	MIT	A	47	Claude Code, Codex
spark-engineerJeffallan/claude-skills	Skill	12k	7d ago	MIT	A	75	Claude Code
rocketride-debugging-pipelinesrocketride-org/rocketride-server	Skill	20k	yesterday	MIT	A	67	Claude Code, Codex
autorag-doctorMarker-Inc-Korea/AutoRAG	Skill	5.3k	yesterday	Apache-2.0	A	144	Claude Code, Codex