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 skills/purpleailab/decepticon/patchworknpx skills add PurpleAILAB/Decepticon --skill patchworkgit clone --depth 1 https://github.com/PurpleAILAB/DecepticonWhat 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 | $0.00052 | $0.06513 |
| Opus 5 | $0.00026 | $0.03256 |
| Sonnet 5 | $0.00010 | $0.01303 |
| Haiku 4.5 | $0.00005 | $0.00651 |
Grade A, and why
patchwork-dropping-elephant 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.
How it starts
The opening of the file, as written. The whole thing — 195 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Patchwork (Dropping Elephant, Chinastrats, MONSOON, Hangover Group) — Adversary Emulation Profile
Patchwork (MITRE ATT&CK G0040) is a cyber-espionage group believed to operate from India, active since at least 2009 and first publicly documented in December 2015. The group earned its name because much of the code used in its tooling is copied and pasted from online forums and public repositories — a "patchwork" of borrowed exploit code and open-source RATs stitched together with custom malware. Despite this low-sophistication approach, Patchwork compensates with high-quality social engineering and aggressive targeting of diplomatic, government, defense, and research organizations primarily across South Asia — with heavy emphasis on Pakistan and China's foreign relations apparatus. The group's malware arsenal includes the custom BADNEWS RAT (and its Ragnatela variant), the public QuasarRAT, BackConfig, VajraSpy (Android), and multiple document exploit chains leveraging CVE-2017-11882, CVE-2017-0199, CVE-2012-0158, and EPS vulnerabilities.
Attribution & motivation
- Sponsor / nation: India (suspected state-sponsored). Circumstantial evidence — Indian keyboard layouts, IP geolocation, infrastructure registration patterns, and targeting profile — consistently points to a pro-Indian or Indian state entity. Norman Shark's 2013 Operation Hangover report traced infrastructure ownership to Indian firms. No formal government attribution has been issued, but multiple vendors (Kaspersky, Symantec, Palo Alto Unit 42, Trend Micro, Malwarebytes, ESET, Cymmetria, Forcepoint, Volexity) converge on an Indian nexus.
- Motivation: Primarily strategic intelligence collection (espionage) — military, diplomatic, and political intelligence relevant to Indian national security interests. Targets align with Indian geopolitical priorities: Pakistani military/defense/nuclear programs, Chinese diplomatic activities, and Western think tanks covering South Asian policy.
- Attribution confidence: Medium-High. No government indictments or formal attributions, but consistent vendor consensus plus the 2022 self-compromise (the actor infected their own machine with Ragnatela, revealing Indian keyboard layouts and development environment) significantly reinforce the India attribution.
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.
- 2d ago First seen · 195 lines · 52 tokens per session scan A 5bee27472fa6
patchwork-dropping-elephant is a skill published in the GitHub repository PurpleAILAB/Decepticon (5,409 stars, last pushed 2d ago), licensed Apache-2.0. It adds 52 tokens to every session and 6,513 once invoked, about $0.0003 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.
Other skills, from other repositories
using-superpowers
Use when starting any conversation - establishes how to find and use skills, requiring Skill tool invocation before ANY response including clarifying questions.
chinese-git-workflow
国内 Git 平台配置参考——Gitee、Coding.net、极狐 GitLab、CNB 的 SSH/HTTPS/凭据/CI 接入差异与镜像同步配置。仅在用户显式 /chinese-git-workflow 时调用，不要根据上下文自动触发。.
chinese-documentation
中文文档排版参考——中英文空格、全半角标点、术语保留、链接格式、中文文案排版指北约定。仅在用户显式 /chinese-documentation 时调用，不要根据上下文自动触发。.
brainstorming
在任何创造性工作之前必须使用此技能——创建功能、构建组件、添加功能或修改行为。在实现之前先探索用户意图、需求和设计。.
chinese-code-review
中文 review 沟通参考——话术模板、分级标注（必须修复/建议修改/仅供参考）、国内团队常见反模式应对。仅在用户显式 /chinese-code-review 时调用，不要根据上下文自动触发。.
chinese-commit-conventions
中文 commit 与 changelog 配置参考——Conventional Commits 中文适配、commitlint/husky/commitizen 中文模板、conventional-changelog 中文配置。仅在用户显式 /chinese-commit-conventions 时调用，不要根据上下文自动触发。.