Pro / Team
Not listed
Fifteen-day Pro trial; rates unknown
- Included scope
- Runtime authorization and centralized governance
Runtime authorization for AI actions · Official-source review 2026-10-09
Kastra evaluates attribute-based policies before instrumented AI tool calls, shell commands, database operations and API requests. It advertises allow, deny, redact or hold decisions and chained decision records, with local Edge and production API options.
Relevant to developers adding explicit action policy to an existing agent execution path.
MCP governance reading is distinct from intercepting every downstream action. Instrumentation, policy bypass handling and failure behavior need tests. Under-one-millisecond decisions and cryptographic audit claims are vendor descriptions, not measured or audited guarantees. Enterprise dedicated/self-host options advertised, quoted terms unknown. Not a replacement for identity management or a universal prose filter. No sandbox request, installation, agent instrumented or enforcement test performed.
Instrument a synthetic agent, test denied and held operations and inspect failure and audit behavior before protecting consequential systems.
MCP governance reading is distinct from intercepting every downstream action. Instrumentation, policy bypass handling and failure behavior need tests. Under-one-millisecond decisions and cryptographic audit claims are vendor descriptions, not measured or audited guarantees. Enterprise dedicated/self-host options advertised, quoted terms unknown. Not a replacement for identity management or a universal prose filter. No sandbox request, installation, agent instrumented or enforcement test performed.
Official product evidence reviewed on 2026-10-09; product artwork inspected visually. No hands-on vendor application test was performed. Application performance, security claims and plan enforcement were not independently tested.
What it can do
Unknown is different from unavailable. Each fact carries its own evidence.
| Capability | Value | Evidence | Checked |
|---|---|---|---|
| Product workflow | Kastra evaluates attribute-based policies before instrumented AI tool calls, shell commands, database operations and API requests. It advertises allow, deny, redact or hold decisions and chained decision records, with local Edge and production API options. | Facts sourced | 2026-10-09 |
| Pricing and restrictions | MCP governance reading is distinct from intercepting every downstream action. Instrumentation, policy bypass handling and failure behavior need tests. Under-one-millisecond decisions and cryptographic audit claims are vendor descriptions, not measured or audited guarantees. Enterprise dedicated/self-host options advertised, quoted terms unknown. Not a replacement for identity management or a universal prose filter. No sandbox request, installation, agent instrumented or enforcement test performed. | Facts sourced | 2026-10-09 |
Understand the total cost
Pro / Team
Not listed
Fifteen-day Pro trial; rates unknown
Enterprise
Not listed
Dedicated/self-host deployment quote
Execlave checks governed agent actions against configured policies before execution and can allow, hold or block a request. It combines a registry, approval controls, spend limits and hash-chained audit evidence in managed or licensed self-hosted deployment.
Explore toolArcjet supplies application and agent protections including bot detection, rate limits, sensitive-data checks and prompt-injection detection. SDKs, agent guards and MCP tooling let developers apply checks around requests and tool actions.
Explore toolCopilotKit connects agents to product interfaces through AG-UI, chat and generative UI components, frontend actions and shared state. Developers can use ready-made components or headless APIs, with hosting and enterprise deployment options.
Explore toolThe practical questions
No. This listing is based on dated official-source evidence and a visual artwork check. Unconfirmed costs and restrictions are identified explicitly.
Kastra evaluates attribute-based policies before instrumented AI tool calls, shell commands, database operations and API requests. It advertises allow, deny, redact or hold decisions and chained decision records, with local Edge and production API options.
No-signup sandbox and fifteen-day Pro trial advertised. Ongoing free tier, trial card requirements and Pro/Team prices unconfirmed.. This record does not confirm an ongoing free plan.
A listed monthly price has not been confirmed. See the plan cards for entitlements, billing commitments and seat minimums.
HTTP API and OpenAPI specification, local Edge integrations and MCP governance-data access advertised. Exact enforcement coverage requires technical documentation.. API access and subscription access may have different terms; consult the linked sources.
MCP governance reading is distinct from intercepting every downstream action. Instrumentation, policy bypass handling and failure behavior need tests. Under-one-millisecond decisions and cryptographic audit claims are vendor descriptions, not measured or audited guarantees. Enterprise dedicated/self-host options advertised, quoted terms unknown. Not a replacement for identity management or a universal prose filter. No sandbox request, installation, agent instrumented or enforcement test performed.
No retained pricing changes yet. A current price alone does not establish a historical trend.
Current official product source read individually; product artwork inspected visually.
Read original source ↗Official pricing and billing restrictions read individually.
Read original source ↗Facts apply to the named version and check date. Send a sourced correction if something changed.