Wlojas Translate
oigfpjciikeohdlpddnekohneanpjfnk
Risk Score
4.24
Risk Level:
Medium
Recommendation:
🟡 MEDIUM RISK — review
Top Risks
- Privacy policy URL unreachable (ConnectTimeout) — no verifiable data handling disclosure for an AI extension processing WhatsApp chat content.
- WhatsApp messages read by content script and potentially sent to three distinct AI provider APIs (Anthropic, OpenAI, Google) with no auditable privacy policy.
- Brand impersonation: extension references WhatsApp but developer (wlojas.com) is not affiliated with Meta/WhatsApp.
- AI/Gen-AI extension processing page content raises exfiltration concern; no CSP present on MV3 extension with innerHTML sink.
- Very low install count (7) with multi-AI-API access creates tail-attack-surface; unverified publisher with unverifiable privacy policy.
Evidence
- privacy_policy_fetch_failed api Privacy policy URL returned ConnectTimeout; policy unverifiable — scored as fetched==false (+10.0 Privacy).
- brand_impersonation store brand_mention.is_impersonation==true for WhatsApp; developer wlojas.com is not confirmed owner.
- ai_multi_api_access manifest Host permissions grant access to Anthropic, OpenAI, and Google Gemini APIs — WhatsApp chat content may be sent to all three.
- dom_xss_sink crx innerHTML assigned from variable in assets/client-Bt4lNSbZ.js; no CSP present to mitigate DOM-XSS.
- no_csp manifest content_security_policy is null on MV3; innerHTML sink without CSP elevates code-quality risk.
- unverified_publisher store verified_publisher==false, is_featured_by_google==false; developer identity unverifiable.
- very_low_installs store Only 7 installs; extension is essentially unreviewed with broad AI API access to sensitive messaging content.
- js_external_hosts crx reactjs.org listed as external JS host alongside 3 AI API domains — reactjs.org is a docs site, not a CDN; unusual.
Permissions Breakdown
- storage low Stores translation preferences/settings locally; low risk.
- host: https://web.whatsapp.com/* medium Content script on WhatsApp Web; can read chat content — sensitive.
- host: https://api.anthropic.com/* medium Sends data to Anthropic AI API; chat content may be transmitted.
- host: https://api.openai.com/* medium Sends data to OpenAI API; chat content may be transmitted.
- host: https://generativelanguage.googleapis.com/* medium Sends data to Google Gemini API; chat content may be transmitted.
Pillar Scores
Permissions2.50
Reputation6.00
Network2.00
Webstore4.50
Maintenance0.00
Privacy10.00
Code Quality0.50
CVE Exposure0.00
Bookkeeping
Rubric v3.6
Scored at 2026-08-31 10:15
Listing SHA
7690217e92d3…
Force block
— not fired
Score recovered
no
Elapsed
—