Vikas Gupta

DreamyDetective's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

liked a model 11 days ago

deepseek-ai/DeepSeek-V4-Pro

Text Generation • 862B • Updated 2 days ago • 1.06M • • 3.73k

updated a Space 18 days ago

Paused

MCP

LLM02 — Secured RAG Knowledge Assistant

🔒

LLM02 mitigated — classified retrieval, PII redaction.

updated a Space 19 days ago

Running

MCP

LLM01 — Secured Support Chat (Prompt Injection Mitigations)

🛡

LLM01 mitigated — safe prompt, checks, output redaction.

updated a Space 20 days ago

Runtime error

MCP

LLM01 — Vulnerable Support Chat (Prompt Injection)

🔓

LLM01 demo — secrets in prompt, no input/output guards.

upvoted a changelog 21 days ago

view changelog

Hugging Face Changelog

Spaces agents.md for your coding agents

21 days ago

• 242

updated a Space 21 days ago

Paused

MCP

LLM09 — Secured Advisor (RAG + Citations + Disclaimers)

📚

LLM09 mitigated — verified KB, citations, disclaimers.

published a Space 21 days ago

Paused

MCP

LLM09 — Secured Advisor (RAG + Citations + Disclaimers)

📚

LLM09 mitigated — verified KB, citations, disclaimers.

updated a Space 21 days ago

Paused

MCP

LLM09 — Vulnerable “Expert” Advisor (Misinformation)

💬

LLM09 demo — ungrounded answers, high temperature.

published a Space 21 days ago

Paused

MCP

LLM09 — Vulnerable “Expert” Advisor (Misinformation)

💬

LLM09 demo — ungrounded answers, high temperature.

updated a Space 21 days ago

Paused

MCP

LLM06 — Secured Agent (Least Privilege + Human-in-the-Loop)

✅

LLM06 mitigated — minimal tools, read-only, review queue.

published a Space 21 days ago

Paused

MCP

LLM06 — Secured Agent (Least Privilege + Human-in-the-Loop)

✅

LLM06 mitigated — minimal tools, read-only, review queue.

updated a Space 21 days ago

Paused

MCP

LLM06 — Vulnerable Tool-Calling Agent (Excessive Agency)

⚠

LLM06 demo — overpowered tools, auto-execute, SQLite.

published 2 Spaces 21 days ago

Paused

MCP

LLM06 — Vulnerable Tool-Calling Agent (Excessive Agency)

⚠

LLM06 demo — overpowered tools, auto-execute, SQLite.

Paused

MCP

LLM02 — Secured RAG Knowledge Assistant

🔒

LLM02 mitigated — classified retrieval, PII redaction.

updated a Space 21 days ago

Paused

MCP

LLM02 — Vulnerable RAG Knowledge Assistant

📂

LLM02 demo — sensitive RAG chunks, no RBAC or filters.

published 3 Spaces 21 days ago

Paused

MCP

LLM02 — Vulnerable RAG Knowledge Assistant

📂

LLM02 demo — sensitive RAG chunks, no RBAC or filters.

Running

MCP

LLM01 — Secured Support Chat (Prompt Injection Mitigations)

🛡

LLM01 mitigated — safe prompt, checks, output redaction.

Runtime error

MCP

LLM01 — Vulnerable Support Chat (Prompt Injection)

🔓

LLM01 demo — secrets in prompt, no input/output guards.

updated a Space 21 days ago

Paused

Kong AI Gateway Data Plane

🛡

published a Space 21 days ago

Paused

Kong AI Gateway Data Plane

🛡