Models included
Paid plans include a monthly model usage budget; bring no separate model provider API key for RedCell inference.
AI penetration testing, from your shell
The AI pentesting agent for your terminal
RedCell automates reconnaissance, CVE scanning, and vulnerability verification for authorized security testing—and blocks actions outside your approved scope. Install in one line and run your first scoped assessment from the terminal.
curl -fsSL https://redcell.sh/install.sh | shThen run redcell login and redcell.
macOS, Linux, and Windows · Models included in paid plan budgets · See current plans
Built with
AI penetration testing for authorized security work.
Services used to run RedCell’s product and account workflows.
One connected workflow
Automated penetration testing, from scope to report
RedCell carries an authorized security assessment through discovery, validation, and reporting, with scope and engagement context attached at each stage.
Stage 01 / Discover
See the surface before you test it.
Start with authorized recon for a security audit or cyber defense review. RedCell maps the exposed services and assets that belong in the engagement.
- Passive recon, OSINT, and subdomain discovery
- Port and service fingerprinting
- Findings organized in the engagement
redcell — sample engagement
scope checked❯ redcell "map the approved application scope"
›scope check target.com approved
›host map api.target.com discovered
›service HTTPS / 443 fingerprinted
✓evidence 03 assets saved locally
Example output · target.com
Stage 02 / Validate
Turn scan output into evidence.
Run an authorized security audit with vulnerability checks and exploit testing through proof-of-concept verification, while honoring engagement scope and rules.
- Template-based CVE scanning
- Scoped exploit testing with proof-of-concept verification
- Human review of findings and impact
redcell — sample engagement
scope checked❯ redcell "verify scoped vulnerability findings"
›scope check target.com approved
›template scan known CVEs complete
›verification evidence collected recorded
✓review state operator review ready
Example output · target.com
Stage 03 / Report
Hand off work people can act on.
Turn security audit and exploit testing results into evidence-backed findings for remediation, technical review, or an executive readout.
- Evidence linked to each finding
- Markdown and JSON exports
- A clear handoff for remediation
redcell — sample engagement
scope checked❯ redcell "prepare the engagement report"
›engagement target.com
›findings organized by severity
›evidence linked to each item
✓export Markdown + JSON
Example output · target.com
Why RedCell
Agentic pen testing without the setup tax
RedCell brings model access, scope guardrails, evidence, and extensible tools into one terminal workflow for authorized security work.
Scope guardrails
Define targets and rules of engagement. RedCell blocks prohibited out-of-scope activity.
Evidence with findings
Keep findings organized with supporting evidence and export Markdown or JSON reports.
Your tools, connected
Extend workflows with custom MCP servers, Python scripts, and native security binaries.
Model Context Protocol
Connect MCP tools with Redcell
Connect your tools with MCP
Connect MCP servers you choose and bring their tools into your RedCell workflow. Keep your specialist tools in reach from the terminal, alongside RedCell’s built-in capabilities.
01
Connect
Register and manage MCP servers from the CLI.
02
Discover
See the tools your connected servers make available.
03
Compose
Use those tools as part of your terminal workflow.
$ redcell mcp ...ready to connect
Common MCP examples
choose your serversGitHub
Repos & pull requests
Filesystem
Files & folders
Playwright
Browser automation
Fetch
Web content
RedCell CLI
Tools in the workflow
DiscoverCall toolsReview results
The RedCell advantage
Agentic pentesting with the operator in control
RedCell is built to do more of the repetitive work while keeping authorization, review, and evidence visible to the person running the engagement.
01 / Guardrails
Scope is part of the workflow—not a footnote.
Confirm the authorized targets and rules of engagement before activity begins. Out-of-scope actions, rate-limit violations, and unwanted impact are blocked.
target.com · approvedother targets · blocked
02 / Operator workflow
Work where your tools already live.
Run RedCell from the terminal, connect the tools you rely on, and keep the engagement close to your existing process.
03 / Evidence
Keep the proof with the finding.
Review collected evidence and export findings in Markdown or JSON.
04 / Extensible
Make it work with your stack.
Bring custom MCP servers, Python scripts, and native security binaries into the workflow.
The RedCell platform
AI security testing for every authorized assessment
Move from authorized targets to verified, review-ready findings in one connected workflow.
Authorize the engagement
Assess the attack surface
Validate and report
Scope Guard
Keep every action inside the approved scope.
Read about scope-guarded testingRecon & OSINT
Run authorized recon to map assets, services, and exposure.
See attack surface mappingSecurity validation
Support cyber defense with security audits and scoped exploit testing.
How CVE scanning worksEvidence & reports
Keep proof with findings and export clear reports.
View report formatsAction layer
Scope-aware runs
Tool orchestration
Operator review
Report export
Agent intelligence
Target context
Workflow planning
Tool selection
Finding verification
Evidence mapping
Authorized surface
DomainsWeb applicationsAPIsHosts & servicesCloud assetsCodebasesCustom MCP tools
Made for security teams
AI pentesting for bug bounty, red teams, and AppSec
Move between reconnaissance, assessment, and defensive work without changing the operator’s home base.
Explore RedCell docs01 / 06
Bug bounty reconnaissance
Turn an authorized program scope into an organized view of assets, services, and leads to review.
Explore bug bounty workflows02 / 06
Red team engagements
Support authorized red team assessments with repeatable reconnaissance, scoped testing, and reviewable evidence.
Explore red team workflows03 / 06
Continuous pentesting
Repeat scoped assessments as your applications change.
See continuous pentesting04 / 06
API security testing
Assess authorized APIs as part of a broader security workflow.
Explore API security testing05 / 06
Compliance readiness
Organize evidence-backed findings for security review and remediation.
Explore compliance workflows06 / 06
PtaaS alternative
Run terminal-based assessments with a monthly plan and usage budget.
See the PtaaS alternativeCompare tools
How RedCell compares to other AI pentesting tools
Read side-by-side guides to understand how different security tools fit different workflows.
From the RedCell blog
Learn AI penetration testing
AI pentesting FAQ
Find quick answers to common questions about the platform, pricing, and security.
RedCell·Support & Docs
Still have questions?
Explore the CLI guides or reach out to our team for assistance.
RedCell is an AI security agent that runs in your terminal. It helps with authorized security testing by automating reconnaissance, CVE scanning, vulnerability verification, and evidence-backed reporting. You define the targets and rules of engagement, and RedCell applies scope guardrails during the workflow.
Your next authorized assessment
Run your next authorized assessment from the terminal
Install RedCell, define your scope, and review verified findings with their evidence—all from your terminal.