← Agent Lab
BRAMHA Agent Lab
Tool Registry
BRAMHA decides which tools to run based on your request. Here's what's under the hood.
PL
Playwright 1.x
Integrated
Risk: low
Browser automation & UI testing
Input: URL / project
Output: Screenshots, traces, test results, DOM snapshots
✓ Local
License: Apache-2.0
Browser launch
Screenshot
Video
Tracing
DOM snapshot
Network capture
Multi-browser
Mobile viewport
LI
Google Lighthouse 12.x
Integrated
Risk: low
Performance, SEO & accessibility audits
Input: URL
Output: JSON/HTML audit reports, scores
✓ Local
⚡ Network
License: Apache-2.0
Performance score
SEO audit
Accessibility
Best practices
PWA check
Core Web Vitals
AX
axe-core 4.x
Integrated
Risk: low
WCAG accessibility testing engine
Input: DOM / URL
Output: Accessibility violations, incomplete, passes
✓ Local
License: MPL-2.0
WCAG 2.1 AA
WCAG 2.1 AAA
Color contrast
ARIA validation
Form labels
Keyboard navigation
SI
sitespeed.io 34.x
Integrated
Risk: low
Deep performance analysis & Core Web Vitals
Input: URL
Output: HAR waterfall, metrics, video, trends
✓ Local
⚡ Network
License: MIT
Core Web Vitals
HAR waterfall
Loading video
Performance trends
Multiple iterations
Mobile testing
ST
Strix latest
Integrated
Risk: high
Agentic security testing platform
Input: Authorised URL / project
Output: Security findings with evidence
✓ Local
⚡ Network
🔒 Sandbox
License: Apache-2.0
Vulnerability scanning
Agentic exploration
Plan & investigate
Validate findings
Local/self-hosted
Ollama/vLLM support
TR
Trivy 0.x
Integrated
Risk: low
Dependency scanning, secrets, SBOM
Input: Filesystem / repository / container
Output: Vulnerability list, misconfigs, secrets
✓ Local
License: Apache-2.0
Vulnerability scan
Misconfiguration
Secret detection
SBOM generation
Container scan
Repository scan
BR
Browser Use latest
Planned
Risk: medium
AI-driven browser exploration
Input: URL + task description
Output: Actions, screenshots, extracted data
✓ Local
⚡ Network
License: MIT
AI agent control
Intelligent exploration
Form interaction
Data extraction
Multi-step flows
SW
SWE-agent latest
Planned
Risk: medium
AI agent for code modification
Input: Repository + issue
Output: Code patches, test results
✓ Local
🔒 Sandbox
License: MIT
Repository navigation
Code modification
Test execution
Issue resolution
How BRAMHA uses tools
You say what you want. BRAMHA picks the right tools.
"Test this website completely."
Playwright + Lighthouse + axe-core + sitespeed.io + Strix + Trivy
"Why is login broken on mobile?"
Playwright (mobile viewport) + Console + Network + DOM
"Check my dependencies for vulnerabilities."
Trivy (filesystem scan)
"Is my site accessible?"
axe-core + Lighthouse accessibility + Playwright keyboard nav