hunt-llm-ai
elementalsouls/Claude-BugHunterHunt LLM/AI feature bugs — prompt injection, indirect injection, exfiltration via tool-use/markdown, ASCII smuggling, agentic AI security (OWASP Agentic Apps…
Scores out of 100 · grade A
2026-08-17Works
40% of the score100/100
- Loads cleanly: valid frontmatter, required fields present, no dangling references.
Maintained
25% of the score100/100
- no commits in the last 12 weeks
Adopted
20% of the score49/100
- 3,628 stars on the source repo.
Documented
15% of the score81/100
- 2,128 words with worked examples.
Install
npx skills add elementalsouls/Claude-BugHunter/hunt-llm-aiWhat it says it does
Hunt LLM/AI feature bugs — prompt injection, indirect injection, exfiltration via tool-use/markdown, ASCII smuggling, agentic AI security (OWASP Agentic Apps 2026, ASI01-ASI10). Patterns: direct injection ('ignore previous instructions'), indirect injection via documents/web pages/email the model reads, ASCII smuggling (Unicode Tags block U+E0000-U+E007F, invisible to humans, decoded by the model), tool-use exfiltration (model has fetch/browse tool, attacker injects OOB URL, model exfils chat history/secrets), markdown-image zero-click exfil, system-prompt extraction, IDOR-via-AI (cross-tenant data). Targets: chatbots, RAG, summarizers, agentic copilots, MCP tools. Detection: any LLM-backed endpoint, doc upload triggering AI processing, autonomous agent with tools. Validate: OOB/Collaborator callback for exfil, verbatim-reproducible system-prompt leak (run twice), verifiable cross-tenant leak or RCE. Confabulation is NOT a finding. Use when hunting AI features, chatbots, RAG, agentic systems, MCP.
Also in elementalsouls/Claude-BugHunter
| Artifact | Score | What the check found | Type | Reach | Last commit |
|---|---|---|---|---|---|
| apk-redteam-pipelineelementalsouls/Claude-BugHunter | clean | Skill | 3.6k stars | today | |
| bb-methodologyelementalsouls/Claude-BugHunter | clean | Skill | 3.6k stars | today | |
| bugcrowd-reportingelementalsouls/Claude-BugHunter | clean | Skill | 3.6k stars | today | |
| cloud-iam-deepelementalsouls/Claude-BugHunter | clean | Skill | 3.6k stars | today | |
| enterprise-vpn-attackelementalsouls/Claude-BugHunter | clean | Skill | 3.6k stars | today | |
| evidence-hygieneelementalsouls/Claude-BugHunter | clean | Skill | 3.6k stars | today |
Other ai & agents skills
Browse all| Artifact | Score | What the check found | Category | Reach | Last commit |
|---|---|---|---|---|---|
| mcp-builderanthropics/skills | No license | AI & agents | 104,949 installs | today | |
| agent-developmentanthropics/claude-plugins-official | clean | AI & agents | 5,901 installs | today | |
| plugin-settingsanthropics/claude-plugins-official | clean | AI & agents | 5,465 installs | today | |
| skill-creatoranthropics/claude-plugins-official | clean | AI & agents | 5,836 installs | today | |
| microsoft-foundrymicrosoft/azure-skills | clean | AI & agents | 544,819 installs | today | |
| claude-apianthropics/skills | No license | AI & agents | 59,236 installs | today |
Put this measurement in your README
A badge carrying how many listings this index holds from the repository and how many pass every static structural check. It reads from this index every time somebody loads your page, so it changes when the measurement changes and there is nothing to keep up to date. Free, no account, and the value is not something you or we can set by hand.
[](https://skillworks.kynth.studio/?q=elementalsouls%2FClaude-BugHunter)Would rather not hotlink us? Every badge is also served in shields.io’s endpoint schema, so shields renders the image and your readers never talk to our domain:
Published by Toolproof, the masthead over this index and eight others. The method behind the number is at toolproof.kynth.studio/methodology, and the whole thing is readable as JSON with no key at /api.
