ZeroLeaks' mission is to make AI agents safe to trust with real tools and data.
Joined March 2025
- Tweets8
- Following2
- Followers2.6K
- Likes371
Pinned Tweet
Introducing Shield 1.1.
Shield checks the text your agent reads (tool results, web pages, MCP tool descriptions, skill files) for prompt injections and jailbreaks before the agent acts on it.
In 1.1, Shield Large catches more attacks and flags fewer benign texts: 4.1%, down from 5.4%, and lower on all six false-alarm datasets. Prompt templates went from 7.0% to 2.3%, coding-agent sessions from 5.3% to 2.2%.
zeroleaks.ai/blog/shield-1-1
teapawt.org/lucknite/status/211585…
Introducing Shield 1.1
Shield is the most accurate detector for prompt injections and jailbreaks. AI agents read text they didn't write: web pages, tool results, MCP tool descriptions, skill files.
Any of it can carry instructions that hijack the agent. Your app sends that text to Shield before the agent acts on it, and Shield tells you whether it's an attack. One API call.
Shield 1.1 makes Shield Large, our most accurate model, better.
ZeroLeaks retweeted
Introducing Shield 1.1.
Shield checks the text your agent reads (tool results, web pages, MCP tool descriptions, skill files) for prompt injections and jailbreaks before the agent acts on it.
In 1.1, Shield Large catches more attacks and flags fewer benign texts: 4.1%, down from 5.4%, and lower on all six false-alarm datasets. Prompt templates went from 7.0% to 2.3%, coding-agent sessions from 5.3% to 2.2%.
zeroleaks.ai/blog/shield-1-1
teapawt.org/lucknite/status/211585…
Introducing Shield 1.1
Shield is the most accurate detector for prompt injections and jailbreaks. AI agents read text they didn't write: web pages, tool results, MCP tool descriptions, skill files.
Any of it can carry instructions that hijack the agent. Your app sends that text to Shield before the agent acts on it, and Shield tells you whether it's an attack. One API call.
Shield 1.1 makes Shield Large, our most accurate model, better.
ZeroLeaks retweeted
Introducing Shield 1.1
Shield is the most accurate detector for prompt injections and jailbreaks. AI agents read text they didn't write: web pages, tool results, MCP tool descriptions, skill files.
Any of it can carry instructions that hijack the agent. Your app sends that text to Shield before the agent acts on it, and Shield tells you whether it's an attack. One API call.
Shield 1.1 makes Shield Large, our most accurate model, better.
ZeroLeaks retweeted
Shield is live.
Prompt injection and jailbreak detection, with four hosted tiers and an OpenAI-compatible API.
Small weights are available for noncommercial use. The small API tier is free.
Launch post: zeroleaks.ai/blog/shield
Try it: zeroleaks.ai/shield
teapawt.org/Lucknite/status/210606…
Introducing Shield: prompt injection and jailbreak detection.
Shield Large ranks #1 overall in our 35-dataset comparison, ahead of GPT-6 Sol, Jev and OpenAI Guardrails.
90.4% mean balanced accuracy. 38ms median model inference on one A10G.
Small weights + free API tier.
ZeroLeaks retweeted
The Small model is fully free plus open weights, Base and Large will be available for free within 2 weeks
Introducing Shield: prompt injection and jailbreak detection.
Shield Large ranks #1 overall in our 35-dataset comparison, ahead of GPT-6 Sol, Jev and OpenAI Guardrails.
90.4% mean balanced accuracy. 38ms median model inference on one A10G.
Small weights + free API tier.
ZeroLeaks retweeted
Introducing Shield: prompt injection and jailbreak detection.
Shield Large ranks #1 overall in our 35-dataset comparison, ahead of GPT-6 Sol, Jev and OpenAI Guardrails.
90.4% mean balanced accuracy. 38ms median model inference on one A10G.
Small weights + free API tier.
Shield is live.
Prompt injection and jailbreak detection, with four hosted tiers and an OpenAI-compatible API.
Small weights are available for noncommercial use. The small API tier is free.
Launch post: zeroleaks.ai/blog/shield
Try it: zeroleaks.ai/shield
teapawt.org/Lucknite/status/210606…
Introducing Shield: prompt injection and jailbreak detection.
Shield Large ranks #1 overall in our 35-dataset comparison, ahead of GPT-6 Sol, Jev and OpenAI Guardrails.
90.4% mean balanced accuracy. 38ms median model inference on one A10G.
Small weights + free API tier.