Penter Desktop · AI-native pentesting

Security testing, powered by agents.Controlled by you.

Plan authorized assessments, run real security tools, and turn verified findings into evidence — all in one operator-controlled workspace.

Don't want a desktop?Use our CLI mode · Run penter in CMD / PowerShell
Scope-aware Operator-visible Evidence-linked
Capability registry
35 registered security tools
NmapSubfinderhttpxKatanaffufNucleiSemgrepTrivy
See how PENTER uses them
Built around the engagement

From scope to evidence, without leaving the workspace.

Penter keeps agents, target context, real tools, requests and findings in one operator-controlled flow.

Agents

Hand off the investigation. Keep every action visible.

Penter agents work from the engagement context already on screen: targets, sessions, prior output and findings. You can see what each agent is doing instead of handing work to a black box.

See the agent workflow
Agent workflow
PENTER / FLOW
01 / 03Give a goalEngagement context
02 / 03Delegate stepsAgent coordination
03 / 03Review the runVisible decisions
From a scoped goal to an inspectable result.
Findings canvas

See the engagement exactly as Penter sees it.

Readiness, severity gaps, verification state and reporting progress live in the same findings canvas used in Penter Desktop.

Explore findings
Findings canvas
PENTER / FLOW
01 / 03DiscoverSurface and signals
02 / 03VerifyTrack confidence
03 / 03ReportLinked evidence
A clear path from discovery to verified evidence.
Tools

Security workflows with every request in context.

Scoped Request Lab replays, HTTP response comparisons, Browser Agent evidence and findings stay linked in the same Penter Desktop workspace.

Explore Request Lab
Scoped Request Lab
PENTER / FLOW
01 / 03SelectAuthorized target
02 / 03CompareRequest and response
03 / 03LinkFinding evidence
HTTP requests, response history and evidence stay linked.
Browser agent

Drive the target from the real browser-agent workspace.

The same Browser Agent view from Penter Desktop keeps the live page, network activity, console output and agent timeline together.

See browser agent
Browser agent
PENTER / FLOW
01 / 03NavigateLive target context
02 / 03InspectRequests and events
03 / 03CaptureTraceable evidence
Browser context stays connected to the investigation.
Learn

Understand the system behind the workflow.

Practical guides for teams evaluating AI-assisted pentesting, agent autonomy and how real security tools fit into an engagement.

Built for control

Autonomous when you want it. Always under your control.

Give agents room to investigate without losing oversight. Every authorized run keeps its boundaries, approvals, tool activity and evidence visible in one place.

Scoped execution Human approvals Traceable evidence
01

Bounded by scope

Agent actions stay tied to the authorized engagement.

02

Approve sensitive actions

Review consequential commands before execution.

03

See the entire run

Follow plans, tools and execution state as they happen.

04

Keep the evidence

Preserve linked requests and verified findings for review.

Pricing

One workspace. Scale the cloud around it.

Penter Desktop stays at the center on every plan. Upgrade when you need stronger models, more cloud credits or more agents working in parallel.

Local-first
Free
€0forever

Full Penter Desktop for local-first security work, with a small Cloud allowance included.

150
Cloud AI credits / month
Cheap cloud profile
Model access
Plan includes
4 core benefits
Full Penter Desktop workspace
Agent, Ask & Plan workflows
Browser Agent + findings & evidence
Local models, BYOK + MCP extensions
Heavy autonomous work
Pro+
Highest capacity
€59/ month

For heavier autonomous investigations, more parallel verification and the highest Cloud allowance.

6,000
Cloud AI credits / month
Reasoning-first workflows
Model access
Plan includes
4 core benefits
Everything in Pro
8 concurrent cloud agents
Longer autonomous investigations
Maximum monthly Cloud capacity
Local/BYOK usage does not consume Penter Cloud credits.Paid plans can add overage packs when needed.Authorized security testing only.
FAQ

Questions operators ask before the first run.

The short version: Penter is built around authorized engagements, real tools and operator-visible execution.

01What is PENTER?+
An AI-native pentesting workspace for authorized security testing. Desktop runs the engagement; the web dashboard handles account, cloud AI, devices, tokens, billing and remote Cloud chat.
02Who is it for?+
Security engineers, pentesters and red teams that want agents, real tools, browser context and evidence in one operator-controlled workflow.
03Can PENTER run autonomous workflows?+
Yes. Agent and auto-run workflows can progress through a plan while scope, command execution, findings and evidence remain visible to the operator.
04Can I use my own or local models?+
Yes. Desktop supports local-model and BYOK workflows, while paid plans mainly increase Penter Cloud model access, credits and parallel agent capacity.
05Can I use Penter without the desktop app?+
Yes. A Windows installation with CLI integration lets you run penter from a new CMD or PowerShell window for local Ollama chat, planning and code review. Use penter desktop to launch Electron. CLI does not inherit Desktop scanning permissions.
06Where do I start?+
Create an account, download Penter Desktop for Windows, connect the device, then create an authorized engagement and give the agent a goal.
Start testing

Start authorized testing with PENTER.

Download the desktop workspace, then manage cloud AI usage and devices from the dashboard.