Building on HF

3 6 6

Vikas Gupta

DreamyDetective

AI & ML interests

None yet

Recent Activity

updated a dataset 3 days ago

DreamyDetective/chat-logs

liked a model 14 days ago

deepseek-ai/DeepSeek-V4-Pro

updated a Space 22 days ago

DreamyDetective/secured-rag-assistant-llm02

View all activity

Organizations

DreamyDetective 's Spaces 18

Kong AI Gateway Data Plane

🛡

LLM02 — Secured RAG Knowledge Assistant

🔒

LLM02 mitigated — classified retrieval, PII redaction.

LLM01 — Secured Support Chat (Prompt Injection Mitigations)

🛡

LLM01 mitigated — safe prompt, checks, output redaction.

LLM01 — Vulnerable Support Chat (Prompt Injection)

🔓

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

LLM09 — Secured Advisor (RAG + Citations + Disclaimers)

📚

LLM09 mitigated — verified KB, citations, disclaimers.

LLM09 — Vulnerable “Expert” Advisor (Misinformation)

💬

LLM09 demo — ungrounded answers, high temperature.

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

✅

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

LLM06 — Vulnerable Tool-Calling Agent (Excessive Agency)

⚠

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

LLM02 — Vulnerable RAG Knowledge Assistant

📂

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

Flex Gateway

😻

Send test HTTP requests and see responses

Sequentialthinking Mcp

🏢

Sequential Thinking MCP Server (Gradio + Python)

Weather MCP

📊

News App

🌍

Search latest news articles from DuckDuckGo

AI Assistant (smolagents + MCP)

🤖

Chat with an AI assistant that can use external tools

Open Brain

🧠

Capture and semantically search your notes with local embeddings

Weather Server

🌖

Get weather forecasts for any location instantly

Chat UI

💬

Chat with open‑source AI models

Agentic RAG

🔬

Answer questions using ArXiv research papers