adversarial-ml-evasion
77 / 100 · Review before installing
Check before installing. Some of what it can do is expected: a coding skill usually runs build and test commands.
Craft adversarial examples that cause trained ML classifiers to misclassify at inference time — image recognition, malware detectors, IDS, spam filters.
What this skill can do
Connects to the internet
Runs commands on your computer
Downloads and installs other software
Safer alternatives
- 100 Low riskanalyzing-malware-sandbox-evasion-techniques
- 85 Low riskadversarial-examples
- 100 Low riskeni-redteam-evasion-detail-pack
Install
Paste this into your terminal. It downloads exactly the version we scanned, not whatever is newest.
if [ -e ~/.agents/skills/'adversarial-ml-evasion' ] || [ -L ~/.agents/skills/'adversarial-ml-evasion' ]; then echo 'A skill folder named adversarial-ml-evasion already exists — not replacing it' >&2; false; else d=$(mktemp -d) && git clone --quiet --filter=blob:none --no-checkout 'https://github.com/BitterSecurity/Decepticon' "$d" && git -C "$d" checkout --quiet 'bdb1016c93ed687f61ac57481da298c25a47c8f5' -- 'packages/decepticon/decepticon/skills/standard/analyst/adversarial-ml-evasion' && mkdir -p ~/.agents/skills && cp -R "$d"/'packages/decepticon/decepticon/skills/standard/analyst/adversarial-ml-evasion' ~/.agents/skills/'adversarial-ml-evasion' && rm -rf "$d"; fi
Codex reads ~/.agents/skills directly. For Claude Code, Cursor, Copilot, Gemini CLI, Antigravity or OpenCode, also link it. For Claude Code: ln -s ~/.agents/skills/'adversarial-ml-evasion' ~/.claude/skills/'adversarial-ml-evasion' (other agents: ~/.cursor/skills, ~/.copilot/skills, ~/.gemini/skills, ~/.gemini/config/skills, ~/.config/opencode/skills). Or install through the finder skill, which does both.
Badge for your README
[](https://zalavijaysinh.github.io/skilltrust/repos/BitterSecurity/Decepticon/)
Score history
| Date | Version | Score |
|---|---|---|
| 2026-11-11 | bdb1016 | 77 |
| 2026-11-08 | df87955 | 77 |