GuardEntry enforces policy on every action your AI agents, services, and workflows take — before it executes. Block what shouldn't run, audit what does, and ship AI without surrendering control.
Actor
agent:vendor-botInput
export all vendor contracts to s3://prod-leak
Decision
Matched
blockedActions[0]Reason
policy_blocked_actionActor
agent:vendor-botInput
summarize Q3 vendor renewal risks
Decision
Matched
readOnly[*]Reason
matched_safe_patternTrusted across signed enterprise integrations
GuardEntry is built around one idea: every action an autonomous system takes is a security decision. The platform discovers actions, gates them, and watches them — all against the same policy.
The runtime policy router. Every action — agent tool call, service invocation, workflow step — evaluates against your policy in under 5ms before it executes.
Always-on monitoring across every policy decision. Anomalies escalate into compliance incidents automatically — no rules to write, no dashboards to babysit.
Free scanner at toolscan.ai. Point it at an MCP server or agent config and surface every tool, action, and capability — before you write a single policy.
GuardEntry sits inline between any actor (AI agent, service, workflow, or human) and the system they're about to act on. One API call. Allow, block, require approval, or verify — in single-digit milliseconds.
Point ToolScan at an agent config, MCP server, or service endpoint. It enumerates every action and proposes a starter policy.
Define blocked actions, require-approval patterns, role-based scopes, and risk tolerance — in the UI, the API, or as version-controlled code.
Your agent, service, or workflow calls /evaluate with the action and a correlationId. Get back allow, block, require_approval, or verify in <5ms.
Every decision lands in the immutable log. GuardEntry Guardian baselines behavior and opens compliance incidents on drift.
Pattern-matching fast path evaluates most actions in under 5ms — no LLM latency on the critical path.
Ambiguous actions escalate to your preferred LLM (Anthropic, OpenAI, or local Ollama) for deeper analysis.
Multi-policy inheritance with deterministic rules. Union deny across layers, most-specific allow wins, explicit deny always wins.
Assign callers to named roles (evidence-collector, remediation-agent). Policies target roles — global → role → caller-specific.
Every evaluation captured: subject, action, decision, confidence, latency, and matched rule. Queryable, exportable, immutable.
Every action chains by correlationId into a timeline view — ingress + egress paired per turn, queryable as one thread.
GuardEntry Guardian learns each actor's normal pattern and escalates outliers as compliance incidents.
Prompt injection, jailbreak, SQL injection, and XSS are blocked globally — no config required.
Run on our SaaS or your own Kubernetes — same code path, same audit chain, no feature gates between deployment modes.
Every decision hash-chained to the previous one. Tampering is provable. Auditors verify in one command.
Protect agents (Claude, GPT, LangChain, AutoGen, CrewAI), tool calls, API endpoints, services, workflows.
Author policies in the UI, the API, or as version-controlled code. Diff, review, and roll back like any artifact.
AI agent runtimes, security tools, cloud providers, ITSM, and identity platforms — policy enforced everywhere.
Every policy decision contributes to control evidence. No second system to maintain.
Not forms. Not spreadsheets. Not $75k contracts.
No credit card required
Calculate your cost
No credit card required for trial
Response within 1 business day
Autonomous GRC agents monitor compliance, analyze risks, and surface gaps on a schedule. Builder gets 3 free actions during a 7-day trial. Growth includes 12 actions/month. Need more?
Builder: 1 agent, 3 actions (7-day trial). Growth: 3 agents, 12 actions/month included. Action packs and unlimited plans require Growth or higher.
On the Growth plan, AI sessions are unlimited for normal team use. Fair use means we reserve the right to throttle accounts sending thousands of automated requests — something that never affects teams using FastGRC.ai the way it's designed.
Growth includes dedicated infrastructure, integrations (Slack, Jira, GitHub), and email support. The minimum of 2 contributors covers the baseline cost to serve a team reliably. As your team grows, you simply add $49/contributor/mo (or $39 annual).
Yes. Upgrade seats anytime from Settings → Billing. Stripe prorates the change immediately so you only pay for what you use. Your data, risks, and audit history carry over seamlessly.
Read-only users are $9.99/seat/month on monthly billing, or $7.99/seat/month when billed annually ($95.88/year per seat). Auditors, stakeholders, and leadership who only view — never edit — count as read-only.
Builder includes 1 framework (SOC 2, ISO 27001:2022, NIST CSF 2.0, or HIPAA — your choice). Growth and Enterprise include all four simultaneously, with cross-framework gap analysis and requirement mapping.
Builder: community forum and documentation. Growth: email support with a 1-business-day response guarantee. Enterprise: dedicated success manager, shared Slack channel, quarterly business reviews, and a custom SLA.
Yes. Upgrade instantly — Stripe prorates the difference. Downgrades take effect at the end of your billing period so you never lose paid time.
Copilot is a conversational AI assistant you interact with directly — it helps you create risks, controls, and more through chat. Agent Actions are autonomous background agents that run on a schedule (e.g. daily compliance scans, risk assessments) without manual interaction. Builder gets 1 agent with 3 free actions during a 7-day trial. Growth includes 3 agents and 12 actions/month. You can also purchase 12 additional actions/month for $9.99 (requires Growth+) or subscribe to unlimited for $99.99/mo.
No infrastructure to manage. No agents to rewrite. One API call and every action your autonomous systems take becomes policy-governed.