Netskope Extension Risk

Detail view · rubric v3.6
← Back to catalog

Bless

pljbjcehnhcnofmkdbjolghdcjnmekia
Risk Score
3.54
Risk Level: Low
Recommendation: 🟢 LOW RISK — review
Category AI
Installs 100,000
Rating 3.2
Last updated 2026-05-11 (3 months ago)
Manifest version MV3
CSP present ✅ yes
Developer extension@txlabs.org
Verified publisher ✅ yes
Featured by Google ❌ no
Privacy policy link
Web Store open ↗

Top Risks

  • cookies + scripting across 8 high-value AI/commerce origins (ChatGPT, Claude, Google, Amazon, Meta) — broad session-hijack surface
  • No developer name listed; txlabs.org unverified organization identity despite verified-publisher badge
  • js_external_hosts includes raw.githubusercontent.com and huggingface.co — dynamic model/code fetch risk not confirmed but concerning
  • AI-extension processing content on major platforms with rating 3.2 — below-average user satisfaction may reflect behavioral concerns
  • system.cpu + system.memory access suggests potential resource-usage telemetry (cryptomining profile pattern)

Evidence

  • cookies+scripting on AI/commerce origins manifest cookies perm + scripting + host_permissions span ChatGPT, Claude, Google, Amazon, Meta AI — full session access.
  • verified_publisher=true, no developer_name store Publisher badge present but developer_name is empty string; identity opacity elevated.
  • js_external_hosts: raw.githubusercontent.com, huggingface.co crx Extension references GitHub raw and HuggingFace — potential remote model/code loading endpoints.
  • is_ai_extension + content_scripts on major platforms manifest Content scripts injected into Google, ChatGPT, Grok, Amazon, Meta AI — AI extension reading page content.
  • privacy_policy fully scoped and adequate api fetched=true, scope_extension=true, data_collection=true, retention=true, third_party_sharing=true — comprehensive policy.
  • no code_findings, obfuscation_score=0.0, js_files_scanned=0 crx Code scan returned no findings but 0 files scanned — surface area not confirmed clean.
  • rating 3.2 at 100K installs store Below-average rating at significant install base; review fetch failed (SSLError) so red-flag pattern unconfirmed.
  • cve_findings_raw=[], threat_intel clean api No CVEs, no bad hosts, no monetization/affiliate hits, domain resolves, not throwaway.

Permissions Breakdown

  • storage low Persists local extension data; low standalone risk.
  • background low Keeps service worker active; low risk on MV3.
  • system.cpu low Read-only CPU metrics; could profile user workload.
  • system.memory low Read-only memory info; low direct harm.
  • offscreen low Offscreen document for DOM tasks; contained in MV3.
  • scripting high Can inject JS into matched AI/shopping origins at runtime.
  • tabs medium Reads tab URLs and metadata across sessions.
  • declarativeNetRequest medium Can intercept/block/redirect network requests via rules.
  • cookies high Access to cookies on matched high-value AI/commerce origins.
  • host:chatgpt.com high Full access to ChatGPT sessions including auth cookies.
  • host:*.google.com high Broad Google property access including Search, Gmail, Gemini.
  • host:perplexity.ai high AI assistant session access.
  • host:claude.ai high Anthropic Claude session/cookie access.
  • host:grok.com high xAI Grok session access.
  • host:amazon.com high Commerce/account session on major shopping platform.
  • host:meta.ai + gateway.meta.ai high Meta AI session and API gateway access.

Pillar Scores

Permissions6.50
Reputation4.00
Network3.50
Webstore4.00
Maintenance0.00
Privacy0.00
Code Quality0.00
CVE Exposure0.00

Bookkeeping

Rubric v3.6
Scored at 2026-08-28 23:11
Listing SHA ce9aa5ea78b3…
Force block — not fired
Score recovered no
Elapsed