Top Risks

• Anonymous developer using free webmail with no business identity or website
• Broad host permissions (<all_urls>) combined with script injection capability
• Brand impersonation signals: title mentions ChatGPT, Claude, Gemini without verification
• Privacy policy hosted on GitHub Pages with incomplete data collection disclosures
• Low install base (5K) relative to capability level increases tail-attack risk

Evidence

• developer_identity store Free webmail (gmail) developer with no verified publisher badge or business website
• host_permissions manifest <all_urls> granted for AI sidebar functionality across 4 chat platforms
• brand_mention store Title mentions ChatGPT, Claude, Gemini; confirmed_owner: false; is_impersonation: true
• privacy_policy api Fetched from GitHub Pages; scope_extension: true, data_collection: false, retention: false
• content_scripts manifest Content scripts match 4 AI chat domains (chatgpt.com, claude.ai, gemini.google.com, grok.com)
• code_findings crx No malicious code patterns detected; obfuscation_score: 0.0
• cve_findings crx No known CVEs in bundled libraries
• store_metrics store 5,000 installs, rating 4.6, updated 2 months ago (August 2026)

Permissions Breakdown

• activeTab low Temporary access on user gesture
• storage low Local data persistence
• sidePanel low UI surface only
• scripting medium Arbitrary script injection capability
• declarativeNetRequest medium Network request modification
• contextMenus low UI integration only
• <all_urls> host_permissions high Broad host access for AI processing

Pillar Scores

Bookkeeping