TinaMind - The most powerful AI Assistant!
befflofjcniongenjmbkgkoljhgliihe
Risk Score
3.54
Risk Level:
Low
Recommendation:
🟢 LOW RISK — review
Top Risks
- cookies + <all_urls>: can read session cookies on every site the user visits (×1.2 multiplier applied).
- AI extension with content scripts on all URLs — full page content accessible on every site.
- innerHTML sink in content.js creates DOM-XSS surface across all injected pages.
- developer_name empty; verified publisher badge present but no display name reduces accountability.
- Privacy policy discloses third-party sharing with full data collection — broad enterprise data exposure.
Evidence
- cookies + <all_urls> manifest cookies permission paired with <all_urls> host permission; ×1.2 multiplier applied per rubric.
- AI extension with broad host access store AI category + content_scripts on <all_urls>; +2.5 AI/Gen-AI webstore signal applied.
- verified_publisher + featured store Verified publisher and featured badge confirmed; discounts applied with floor 2.0 on reputation.
- dom_sink_innerhtml_userctrl crx innerHTML assignment in content.js; CSP present so +0.5 (not escalated to +2.0).
- privacy_policy_classification api Policy fetched, scoped to extension, data_collection=true, third_party_sharing=true, retention=true.
- js_external_hosts crx 4 external hosts: api.tinamind.com, tinamind.com, wiseinks.tinamind.com, chrome.google.com — all dev-controlled.
- install_perm_anomaly api has_high_tier_permission=true but small_install_high_perm=false and tail_attack_surface=false.
- no bad_host / affiliate / monetization hits api threat_intel shows empty bad_host_hits, affiliate_hits, monetization_hits; domain resolves and not throwaway.
Permissions Breakdown
- tabs medium Access to tab URLs and titles; combined with <all_urls> raises data exposure risk.
- storage low Local extension data storage; low standalone risk.
- cookies high Can read/write cookies across all sites when paired with <all_urls>; high credential risk.
- declarativeNetRequest medium Can modify/block network requests; no full webRequest but still impactful.
- contextMenus low Adds right-click menu items; minimal standalone risk.
- <all_urls> (host_permission) high Content scripts injected into every page; broad read/modify capability on all sites.
Pillar Scores
Permissions6.24
Reputation2.00
Network2.00
Webstore3.50
Maintenance0.00
Privacy1.00
Code Quality0.50
CVE Exposure0.00
Bookkeeping
Rubric v3.6
Scored at 2026-06-16 07:19
Listing SHA
00ff57c60b82…
Force block
— not fired
Score recovered
no
Elapsed
25.1s