AgentAegis reviews an agent's prompt and tool list, rates the risk of each tool call, and suggests input/output filters plus an EU AI Act mapping - so an agent cannot silently exfiltrate data or take an unguarded action.
No credit card required · Cancel anytime
AgentAegis is an AI-agent guardrail layer that scores system prompts and tool calls for security risks - secret leakage, prompt injection, and unsafe tool use - using rule-based checks plus a model-assisted filter, and returns a cited guardrail report.
Built for founders who need results without a learning curve.
Built for founders who need results without a learning curve.
Built for founders who need results without a learning curve.
Built for founders who need results without a learning curve.
Paste system prompt + tool list.
Rule-based risk scoring, then model-assisted filters.
Export guardrail report with runId + ruleset version.
or generate below.
Social proof uses templates until real, consented testimonials exist. Trusted by [X]+ agent builders and compliance leads.
“PENDING REAL QUOTE — replace with a consented customer quote.”
“PENDING REAL QUOTE — outcome metric as template only ([X]+ hours saved).”
“PENDING REAL QUOTE — no fabricated names or logos.”
Secret leakage in prompts, prompt-injection patterns, and unsafe tool or function calls, mapped to the OWASP Top 10 for LLM Applications.
Pro uses the platform LLM key under fair use; Free and demo mode return a realistic mock report without live calls.
Findings reference LLM01 (prompt injection), LLM02 (data exfiltration), and LLM06 (over-reliance) where relevant.
Yes - Pro exports a report with runId and ruleset version for audit.
No. It is a pre-deployment review layer for agent builders, not a network firewall.
Any agent that exposes a system prompt and tool list (LangChain, AutoGen, or a custom stack).
Yes. Self-serve plans cancel anytime; access continues until period end.
No. Start with email signup, then upgrade when ready.
No for Free/Pro. AI runs are included in your plan (fair use) via our platform key. Enterprise can optionally bring your own OpenAI key (BYOK) — configure it at /settings (key stays server-side only).
Pro includes about 300 workflow runs/month (fair use cost guard ≈ 300 gens on gpt-4o-mini, plus a daily cap). If you hit the fair-use limit, Studio returns HTTP 429 — upgrade, wait for reset, enable Demo mode, or use Enterprise BYOK. We never silently return fake AI success.
Payments are processed by Waffo Pancake (merchant of record).
You receive access confirmation; fulfillment is tracked via webhook + order logs.
Join builders using AgentAegis. Free to try — no card needed.
Trusted · Cancel anytime