Discover Every Plugin Installed to Your AI Agents
Akto now surfaces all plugins installed to AI agents, including the skills, MCP servers, and tools bundled inside each one. Full visibility into what every agent in your org can actually do.
AI Discovery
SaaS Agent Grouping
SaaS-based AI agents now grouped separately under SaaS Agents for a cleaner inventory and faster identification.
AI Discovery
Copilot Studio Architecture View
Map Copilot Studio agents to their connected tools, topics, MCP servers, and knowledge sources in one architecture view.
AI Guardrails
LLM-Based Redaction
Define sensitive data in plain language. Akto masks matching content automatically before requests proceed to the model.
AI Guardrails
Guardrail Change Impact Analysis
Preview how guardrail policy edits affect detections before updating. Replay recent violations against current(saved) and drafted policies.
AI Guardrails
Agent Misconfiguration Detection
Scan settings.json and config.toml across Claude Code, Codex CLI, and Copilot CLI endpoints to flag agents running with bypassed permissions, disabled sandboxes, or hooks turned off.
AI Guardrails
Violation Evidence and Remediation
Triage violations by severity with trigger reasons, matched evidence, and risk scores, then act on built-in response steps and hardening recommendations
AI Guardrails
Skill Evaluations Tab
Review skill evaluation results and malicious skill alerts in a dedicated tab within Guardrail Activity. See which skills failed checks and on which endpoints.
AI Guardrails
Statistical Anomaly Guardrail
Detect behavior drift, abuse spikes, and tool misuse automatically by baselining normal tool usage and error rates per agent, without manual rule sets.
Integrations
Claude Inference Hooks
Akto now acts as the AI security server for Anthropic's Inference Hooks. Every prompt across claude.ai, Cowork, and Claude Code gets an allow/deny verdict from Akto before it reaches the model.
Integrations
Amp Code Hooks
Intercept Amp proxy traffic and route it through Akto. Apply guardrail policies to every request flowing through Amp-based coding agents.
Integrations
Glean Connector
Enforce guardrail policies on every employee interaction with Glean's AI agents and capture full traces of prompts, responses, and tool calls in Akto.
Integrations
Amazon Quick Connector
Import Amazon Quick agent activity into Akto and check employee interactions against your guardrail policies asynchronously.
AI Guardrails
Statistical Anomaly Guardrail
Detect behavior drift, abuse spikes, and tool misuse automatically by baselining normal tool usage and error rates per agent, without manual rule sets.
AI Guardrails
Violation Evidence and Remediation
Triage violations by severity with trigger reasons, matched evidence, and risk scores, then act on built-in response steps and hardening recommendations