
shannon
Run Shannon white-box pentests of web apps and APIs you are authorized to test, from your coding agent, with safety rules built in.
- GitHub stars
- 52K
Install
/plugin install shannon@keygraph-shannonRun it inside Claude Code after adding the marketplace with /plugin marketplace add KeygraphHQ/shannon.
Contents
Discovered from sourceSkills 1
Shannon documentation
Before running anything
Shannon actively exploits the target. It can create, change, and delete data.
- Before the first
startin a conversation, ask the user to confirm two things: they own the target or are authorized to test it, and it is not a production system. Do not infer either from a URL, a repository name, or the target being local. - If the target looks like a production host, say so and do not start.
- Never run
reset. It wipes all scan state. Leave it to the user. - Pass
--yestostoponly when the user asked to stop that scan. - Do not run
setupfrom the agent shell. It is interactive. Ask the user to run it in their own terminal, and never ask them to paste a provider key into chat.
If the user asks to skip one of these, decline that part and explain why.
Documentation
Use the documentation bundled at the plugin root:
- llms.txt — concise overview and documentation index. Start here for a short reference.
- llms-full.txt — full documentation. Consult the relevant sections for detailed setup, usage, safety, configuration, and troubleshooting.
Sponsors
Details
- Marketplace
- KeygraphHQ/shannon
- Source
- plugins/shannon
- Category
- Community