MCPTriage

aig-agent-redteam skill

github:Tencent/AI-Infra-Guard/skills/aig-agent-redteam

Caution

Low-impact capabilities worth knowing about before installing.

Why

Changes AI rules or configuration files. in modules/mutation-attack/data/eval_datasets/ChatGPT-Jailbreak-Prompts.json
Matched an automated check. in modules/mutation-attack/data/eval_datasets/ChatGPT-Jailbreak-Prompts.json
Contains invisible or obfuscated characters. in modules/mutation-attack/operators/ascii_smuggling.md
Contains instructions placed where the agent reads them indirectly. in modules/mutation-attack/operators/fake_system_prefix.md

GitHub archives have no registry checksum; the result is tied to the commit shown.

Scanned version: commit 1c69920ff214a149c0653e522fb20f0c04f52776
Scan date: 2026-11-02
Analysis revision: f48103a381866b92
Integrity: No registry checksum (git archive); commit 1c69920ff214a149c0653e522fb20f0c04f52776; archive sha256 1df12341067389df7978d9ea2818d7deff4e1a6fa6442fec1a9363c8a1c1a164
Listed in: VoltAgent/awesome-agent-skills
Licence: Apache-2.0
Repository: https://github.com/Tencent/AI-Infra-Guard

Automated static analysis by MoorAI's open-source engine. A verdict describes what the package's code can do, not the intent of its authors. Report a mistake →