Automated package-squatting & slopsquatting detection signals across PyPI and npm — these are automated, unconfirmed signals. Use them to find candidates to investigate, but don't automatically assume they're malicious.

Sorted by risk score. Cleared packages hidden by default — add verdict:benign_community to include. Packages the registry has since removed are hidden too — see them on the Taken Down tab or add state:taken_down. Very risky packages require manual human audit. • Tag & Signal Legend →

Search & filter syntax reference
Filter with verdict: (malicious, suspicious, unverified_high_signal, squatted_stub…), ecosystem: (pypi/npm), hook: (install, network, pth, gyp, worm, c2, creds, binary, env, stealer, any — malicious hooks & behaviors), email:, domain:, keyword:, version: (comma = OR within a field, space = AND across fields), age:<6h / age:>90d (combine both for a range), state:live / state:taken_down (registry liveness — the feed defaults to state:live), lead:>7d (days we had a since-removed package flagged while it was still up), or a bare word for a package-name search.
Package Score ▼ Age Users Lines Growth Verdict Targeted Ecosystem Version Author Email domain Signals
60 12d — — —
Unverified Signal
— npm latest Simon Dieny — 🪝 🌐
Package purpose: Agent Visualizer: explore your AI context, skills and MCP configuration in a local dashboard.
Static code signals:
  • CREDENTIAL_PATH_HARVESTING: 'IDE / AI Agent Configuration Hijacking' found in package/dist/scan/assets.js:155 (+1 more occurrence(s) elsewhere)
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/dist/browser/main-QX766DTS.js:5
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in package/dist/scanner.js:13

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 12d — — —
Unverified Signal
— npm latest IdeaboticsAI — 🪝 🌐
Package purpose: Proofwire hub — multi-tenant transparency log, policy registry, approvals, and witness service.
Static code signals:
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/src/app.js:2012 (+1 more occurrence(s) elsewhere)
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in package/src/app.js:140 (+3 more occurrence(s) elsewhere)
  • CREDENTIAL_PATH_HARVESTING: 'Cloud Instance Metadata (AWS/GCP/Azure IMDS)' found in package/src/egress.js:12

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 12d — — —
Unverified Signal
— npm latest — — 🪝 🌐
Package purpose: A draggable desktop pill and menu bar app showing your AI plan usage: Claude Code, Ollama Cloud, and more via providers
Static code signals:
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/src/core/providers/chatgpt.mjs:38 (+4 more occurrence(s) elsewhere)
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in package/src/core/providers/chatgpt.mjs:19 (+4 more occurrence(s) elsewhere)
  • CREDENTIAL_PATH_HARVESTING: 'IDE / AI Agent Configuration Hijacking' found in package/src/core/claude-code.mjs:7

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 12d — — —
Unverified Signal
— npm latest — — 🪝 🌐
Package purpose: Turn Markdown into a DAG diagram
Static code signals:
  • SOURCE_CODE_ENCODED_PAYLOAD: 'atob()' found in package/dist/markdag.core.iife.js:38 (+2 more occurrence(s) elsewhere)
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/dist/markdag.iife.js:1
  • SUSPICIOUS_OBFUSCATION: 'Invisible Unicode Steganography / Bidi Override' found in package/dist/markdag.iife.js:1

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 13d — — —
Unverified Signal
— npm latest AirA2A Team — 🪝 🌐
Package purpose: Zero-dependency Meta-MCP Hub for Claude Desktop, Cursor, & AI Agents. Autonomous capability discovery across 470+ Model Context Protocol servers.
Static code signals:
  • CREDENTIAL_PATH_HARVESTING: 'IDE / AI Agent Configuration Hijacking' found in package/bin/cli.js:33
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/bin/cli.js:223
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in package/bin/cli.js:14

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 13d — — —
Unverified Signal
— npm latest Adaptive Mind — 🪝 🌐
Package purpose: Diagnose and fix inefficient AI coding sessions. Local, private, one command.
Static code signals:
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/public/js/app.js:86 (+1 more occurrence(s) elsewhere)
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in package/src/collectors/claude.js:314 (+6 more occurrence(s) elsewhere)
  • CREDENTIAL_PATH_HARVESTING: 'IDE / AI Agent Configuration Hijacking' found in package/src/suggestions/targets.js:47

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 13d — — —
Unverified Signal
— npm latest Viacheslav Ptsarev — 🪝 🌐
Package purpose: MAX Messenger bots through the official Bot API, and a personal MAX account, from the terminal — built for agents and scripts
Static code signals:
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in package/dist/bot/auth.js:11 (+24 more occurrence(s) elsewhere)
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/dist/download.js:102 (+3 more occurrence(s) elsewhere)
  • CREDENTIAL_PATH_HARVESTING: 'SSH Directory / Private Keys (~/.ssh)' found in package/dist/commands/messages.js:244 (+1 more occurrence(s) elsewhere)

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 15d — — —
Unverified Signal
— npm latest Arthur Sultanbekov — 🪝 🌐
Package purpose: CLI for SoftPractice, an online platform for project-based programming practice
Static code signals:
  • LIFECYCLE_SCRIPT: 'postinstall' -> 'node scripts/install.js'
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/scripts/install.js:16
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'execSync/spawnSync' found in package/bin/softpractice.js:18

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 16d — — —
Unverified Signal
— npm latest — — 🪝 🌐
Static code signals:
  • LIFECYCLE_SCRIPT: 'preinstall' -> 'npx only-allow pnpm'
  • LIFECYCLE_SCRIPT: 'postinstall' -> 'patch-package'
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/dist/Dialogs/CreatePhrase/CreatePhrase.js:411 (+2 more occurrence(s) elsewhere)

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 18d — — —
Very Risky
— pypi 2.0.1 Michał Jaworski — 🪝
Package purpose: Cython-based Python bindings for dear meltygui_imgui
Static code signals:
  • INSTALL_TIME_EXECUTION: 'eval' executed at top-level in meltygui_imgui-2.0.1/setup.py:60
  • SOURCE_CODE_ENV_VARS_ACCESS: Call to os.environ.get() in meltygui_imgui-2.0.1/setup.py:20
  • SOURCE_CODE_ENV_VARS_ACCESS: Access to os.environ in meltygui_imgui-2.0.1/setup.py:20
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'eval()' found in meltygui_imgui-2.0.1/setup.py:60
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in meltygui_imgui-2.0.1/setup.py:20
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'Python Subprocess / OS Execution' found in meltygui_imgui-2.0.1/tools/build.py:41
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'Python Subprocess / OS Execution' found in meltygui_imgui-2.0.1/tools/build_windows.py:40
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in meltygui_imgui-2.0.1/tools/build_windows.py:31

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 18d — — —
Unverified Signal
— npm latest — — 🪝 🌐
Package purpose: A Rust CLI to track your daily accomplishments in Obsidian
Static code signals:
  • LIFECYCLE_SCRIPT: 'postinstall' -> 'node scripts/install.js'
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/scripts/install.js:109
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'execSync/spawnSync' found in package/bin/swelog.js:22

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 18d — — —
Unverified Signal
— npm latest — — 🪝 🌐
Package purpose: Auth0 Node.js SDK for the Management API v2.
Static code signals:
  • SOURCE_CODE_ENCODED_PAYLOAD: 'Buffer.from(..., 'base64')' found in package/dist/cjs/management/core/base64.js:22
  • SOURCE_CODE_ENCODED_PAYLOAD: 'atob()' found in package/dist/cjs/management/core/base64.js:6
  • SUSPICIOUS_OBFUSCATION: 'Layered Base64 and Decompress Pipeline' found in package/dist/cjs/management/core/base64.js:22
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/dist/cjs/management/Client.js:298

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 19d — — —
Unverified Signal
— npm latest 2.13 2.13 🪝 🌐
Package purpose: SuperPrint by 2.13 — professional desktop publishing (DTP) and prepress, in your browser. Free, no subscription, no account. CMYK/PDF up to 600 DPI with bleed & imposition, advanced typography, vector text export, offline PWA, the SP213 AI layout studio (
Static code signals:
  • LIFECYCLE_SCRIPT: 'postinstall' -> 'node scripts/postinstall.mjs'
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'execSync/spawnSync' found in package/cli.mjs:165
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/cli.mjs:375

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 20d — — —
Unverified Signal
— npm latest — — 🪝 🌐
Package purpose: Stateful Linear GraphQL API emulator for local development and CI.
Static code signals:
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/dist/index.js:174
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in package/dist/index.js:2033
  • SUSPICIOUS_OBFUSCATION: 'Layered Base64 and Decompress Pipeline' found in package/dist/index.js:2737

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 21d — — —
Unverified Signal
— npm latest — — 🪝 🌐
Package purpose: Autonomous AI-powered penetration-testing swarm — one command, real vulns.
Static code signals:
  • LIFECYCLE_SCRIPT: 'postinstall' -> 'node install.js'
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'require('child_process')' found in package/bin/cli.js:8
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'execSync/spawnSync' found in package/bin/cli.js:22
  • SOURCE_CODE_NETWORK_CALL: 'require('http(s)')' found in package/install.js:9

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 22d — — —
Unverified Signal
— npm latest — — 🪝 🌐
Package purpose: Claude Code lifecycle contextual hooks for Mnemonik. PreToolUse Edit/Write/NotebookEdit/Bash gates plus SessionStart bootstrap warming, clean Stop, PreCompact context_snapshot creation, bounded SessionEnd cleanup, and UserPromptSubmit context injection. T
Static code signals:
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in package/dist/lib/boundContext.js:35 (+5 more occurrence(s) elsewhere)
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/dist/lib/env.js:117
  • ANTI_ANALYSIS_EVASION: 'CI/CD & Sandbox Environment Detection' found in package/dist/lib/env.js:22
  • CREDENTIAL_PATH_HARVESTING: 'IDE / AI Agent Configuration Hijacking' found in package/dist/install.js:9 (+1 more occurrence(s) elsewhere)

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 24d — — —
Unverified Signal
— npm latest Oluwayemi Oyinlola — 🪝 🌐
Package purpose: Command-line interface for scaffolding, generating, and managing Zudojs framework projects.
Static code signals:
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/dist/src/adapters/frontend/angular.adapter.js:107 (+10 more occurrence(s) elsewhere)
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in package/dist/src/cliApplication/cliApplication.core.js:43 (+11 more occurrence(s) elsewhere)
  • SUSPICIOUS_OBFUSCATION: 'Invisible Unicode Steganography / Bidi Override' found in package/dist/src/cliApplication/invocation/invocation.sanitize.js:12

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 29d — — —
Unverified Signal
— npm latest Kirill Lakhtachev — 🪝 🌐
Package purpose: One seat for all your projects and tasks: a local record of what is done, what is next and when it ships - read by you and your coding assistant
Static code signals:
  • LIFECYCLE_SCRIPT: 'postinstall' -> 'node install.js'
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'require('child_process')' found in package/download.js:7 (+1 more occurrence(s) elsewhere)
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'execSync/spawnSync' found in package/download.js:67 (+1 more occurrence(s) elsewhere)
  • SOURCE_CODE_NETWORK_CALL: 'require('http(s)')' found in package/download.js:5

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 30d — — —
Unverified Signal
— npm latest Jelou Inc. — 🪝
Package purpose: Build AI agents on the Jelou platform — from your terminal or your AI editor.
Static code signals:
  • LIFECYCLE_SCRIPT: 'postinstall' -> 'node bin/postinstall.js'
  • CREDENTIAL_PATH_HARVESTING: 'Registry / Git Credentials (~/.npmrc, ~/.pypirc)' found in package/bin/postinstall.js:45

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 32d — — —
Unverified Signal
— npm latest — — 🪝 🌐
Package purpose: agentop CLI — npm install for the same native binary distributed via install.sh (Linux x86_64 only)
Static code signals:
  • LIFECYCLE_SCRIPT: 'postinstall' -> 'node scripts/postinstall.js'
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'require('child_process')' found in package/bin/agentop.js:8
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'execSync/spawnSync' found in package/bin/agentop.js:21
  • SOURCE_CODE_NETWORK_CALL: 'require('http(s)')' found in package/scripts/postinstall.js:13

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 33d — — —
Unverified Signal
— npm latest Carsen Klock — 🪝 🌐
Package purpose: Self-hosted, open-source AI agent platform — CLI
Static code signals:
  • LIFECYCLE_SCRIPT: 'postinstall' -> 'node scripts/postinstall.cjs'
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'require('child_process')' found in package/bin/cybara.cjs:3
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'execSync/spawnSync' found in package/bin/cybara.cjs:17
  • SOURCE_CODE_NETWORK_CALL: 'require('http(s)')' found in package/scripts/download.cjs:4
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/scripts/download.cjs:27

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 33d — — —
Unverified Signal
— npm latest Jake Runzer — 🪝 🌐
Package purpose: Develop and deploy code with zero configuration
Static code signals:
  • LIFECYCLE_SCRIPT: 'postinstall' -> 'node ./npm-install/postinstall.js'
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/npm-install/postinstall.js:84
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'execSync/spawnSync' found in package/bin/railway.js:14

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 34d — — —
Unverified Signal
— npm latest Ossian — 🪝 🌐
Package purpose: A Node.js HTTP request library based on libcurl, with browser-grade TLS/HTTP2/HTTP3 fingerprint customization. It patches BoringSSL to customize the TLS client hello (cipher suites, extensions, order) and uses a custom HTTP/2 / HTTP/3 implementation, so t
Static code signals:
  • LIFECYCLE_SCRIPT: 'install' -> ':'
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/dist/fetch.js:6 (+2 more occurrence(s) elsewhere)
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'require('child_process')' found in package/scripts/postinstall.js:11
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'execSync/spawnSync' found in package/scripts/postinstall.js:12

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 35d — — —
Unverified Signal
— npm latest — — 🪝 🌐
Package purpose: Kubernetes lexicon for chant — declarative IaC in TypeScript
Static code signals:
  • ANTI_ANALYSIS_EVASION: 'TLS / SSL Verification Bypass (Defense Evasion)' found in package/src/op/activities/argo.ts:76
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/src/op/activities/argo.ts:78
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in package/src/op/activities/argo.ts:75 (+2 more occurrence(s) elsewhere)
  • CREDENTIAL_PATH_HARVESTING: 'Kubernetes Service Account Token' found in package/src/composites/cockroachdb-cluster.ts:425 (+2 more occurrence(s) elsewhere)
  • CREDENTIAL_PATH_HARVESTING: 'Kube / Docker Credentials' found in package/dist/api/fake-cluster.d.ts:8 (+1 more occurrence(s) elsewhere)

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 48d — — —
Unverified Signal
— npm latest — — 🪝 🌐
Package purpose: Four-source migration wizard: copy Claude Code, Codex, OpenCode and Hermes sessions, memories, skills, instructions and slash commands into DeepSeek Harness as resumable sessions — copy-only, approval-gated, idempotent
Static code signals:
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/client/client.js:306
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in package/lib/agmd-section.mjs:24 (+6 more occurrence(s) elsewhere)
  • CREDENTIAL_PATH_HARVESTING: 'SSH Directory / Private Keys (~/.ssh)' found in package/lib/settings.mjs:41

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 58d — — —
Unverified Signal
— npm latest Kirill Lakhtachev — 🪝 🌐
Package purpose: Work activity tracker CLI: automatic workday and break detection, task management with Jira/GitLab integration, productivity reports and exports
Static code signals:
  • LIFECYCLE_SCRIPT: 'postinstall' -> 'node install.js'
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'require('child_process')' found in package/download.js:7 (+1 more occurrence(s) elsewhere)
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'execSync/spawnSync' found in package/download.js:63 (+1 more occurrence(s) elsewhere)
  • SOURCE_CODE_NETWORK_CALL: 'require('http(s)')' found in package/download.js:5

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 103d — — —
Unverified Signal
— npm latest — — 🪝
Package purpose: OS-level Sandbox adapter for @lite-agent/core over @anthropic-ai/sandbox-runtime (Seatbelt / bubblewrap).
Static code signals:
  • CREDENTIAL_PATH_HARVESTING: 'SSH Directory / Private Keys (~/.ssh)' found in package/dist/index.js:76
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'execSync/spawnSync' found in package/dist/index.js:29

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 162d 22 4 ↑0.1/day
SUSPICIOUS
okta npm 45.0.0 — —
Package purpose: Placeholder package
Matched naming pattern: node-okta-app-launcher-okta-b2b
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 162
Code-only verdict: Possible Squatted Stub (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 22/month, 5/week (NEGLIGIBLE_OR_ZERO_USAGE)
Codebase size: 4 lines, 79 Bytes across 1 file (TINY_CODEBASE)
Automated signals flagged:
  • MEDIUMPackage name '@engie-ec/app-launcher-okta-b2b' matches AI hallucination template [node] + [okta] + [app-launcher-okta-b2b].
  • HIGHClaims critical enterprise brand identity ('OKTA').
  • INFOPackage release verified with cryptographic build provenance (npm_slsa_sigstore). Guarantees authentic repository build pipeline and eliminates publisher domain spoofing risk.
  • MEDIUMPackage has empty or near-zero description (potential reservation stub).
  • HIGHPackage registered recently (143 days ago) during the active AI hallucination slopsquatting wave.
  • HIGHPackage registered with suspiciously high major version (v45.0.0, major 45) despite recent registration (143 days ago, 1 release(s)), indicating potential dependency confusion attack to shadow internal organization builds.
  • HIGHMISSING_SOURCE_REPOSITORY_URL
  • INFOPackage affiliated with trusted vendor 'okta' (provenance:npm_slsa_sigstore). Applied 50% threat score dampening (120 -> 60) to reduce false positives while retaining hijack/takeover detection.
Static code signals:
  • MISSING_SOURCE_REPOSITORY_URL

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 187d — — —
Unverified Signal
— npm latest Yuhan Lei — 🪝
Package purpose: Claude Code statusline and rate limit tracker. Pure Bash + jq, single file.
Static code signals:
  • CREDENTIAL_PATH_HARVESTING: 'IDE / AI Agent Configuration Hijacking' found in package/cli.js:51
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'require('child_process')' found in package/cli.js:7
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'execSync/spawnSync' found in package/cli.js:20

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 199d 308 69 ↑1.5/day
SUSPICIOUS
llama pypi 0.1.0 MiniMax — 🪝
Package purpose: llama-index llms minimax integration
Claimed homepage: https://pypi.org/project/llama-index-llms-minimax/
Matched naming pattern: llama-llama-llms
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 199
Code-only verdict: BENIGN_COMMUNITY (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 308/month, 48/week (MODERATE_USAGE)
Codebase size: 69 lines, 2.6 kB across 3 files (TINY_CODEBASE)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('LLAMA').
  • HIGHPublisher email 'None' is not affiliated with official vendor domain.
  • MEDIUMPackage published 5+ versions in rapid succession (< 24h) to mimic mature open source maintenance.
  • MEDIUMPackage registered 190 days ago within the 1-year AI tool proliferation window.
  • HIGHSOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in llama_index_llms_minimax-0.1.0/llama_index/llms/minimax/base.py:43
  • INFOModerate community usage with 308 monthly downloads.
Static code signals:
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in llama_index_llms_minimax-0.1.0/llama_index/llms/minimax/base.py:43

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 209d 318 2,488 ↑1.5/day
SUSPICIOUS
llama pypi 6.0.0 — contact@wiscale.fr wiscale.fr
Package purpose: LlamaIndex VectorStore for VelesDB: The Local AI Memory Database. Microsecond RAG retrieval.
Claimed homepage: https://pypi.org/project/llama-index-vector-stores-velesdb/
Matched naming pattern: llama-llama-vector
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 209
Code-only verdict: BENIGN_COMMUNITY (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 318/month, 43/week (MODERATE_USAGE)
Codebase size: 2488 lines, 115.1 kB across 13 files (LARGE_CODEBASE)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('LLAMA').
  • HIGHPublisher email 'contact@wiscale.fr' is not affiliated with official vendor domain.
  • MEDIUMPackage published 5+ versions in rapid succession (< 24h) to mimic mature open source maintenance.
  • MEDIUMPackage registered 198 days ago within the 1-year AI tool proliferation window.
  • MEDIUMPackage registered with suspiciously high major version (v6.0.0, major 6) despite recent registration (198 days ago, 54 release(s)), indicating potential dependency confusion attack to shadow internal organization builds.
  • INFONo malicious install-time hooks or shell cradles detected in package AST.
  • INFOSubstantial functional codebase (2488 LOC across 13 file(s)).
  • INFOModerate community usage with 318 monthly downloads.

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 222d 143 4 ↑0.6/day
Possible Squatted Stub
gcp pypi 1.0.29 — —
Package purpose: Transform library for GCP data pipelines (dbt macros)
Claimed homepage: https://pypi.org/project/gcp-pipeline-transform/
Matched naming pattern: gcp-gcp-transform
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 222
Registry downloads: 143/month, 44/week (MODERATE_USAGE)
Codebase size: 4 lines, 62 Bytes across 1 file (EMPTY_STUB)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('GCP').
  • HIGHPublisher email 'None' is not affiliated with official vendor domain.
  • MEDIUMPackage registered 212 days ago within the 1-year AI tool proliferation window.
  • INFONo malicious install-time hooks or shell cradles detected in package AST.
  • MEDIUMPackage is an empty placeholder (4 LOC, 1 file(s), 62 bytes).
  • INFOModerate community usage with 143 monthly downloads.

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 225d 143 123 ↑0.6/day
SUSPICIOUS
llama pypi 0.2.0 — you@example.com example.com
Package purpose: llama-index tools igpt_email integration
Claimed homepage: https://pypi.org/project/llama-index-tools-igpt-email/
Matched naming pattern: llama-llama-tools
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 225
Code-only verdict: BENIGN_COMMUNITY (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 143/month, 18/week (MODERATE_USAGE)
Codebase size: 123 lines, 5.2 kB across 2 files (TINY_CODEBASE)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('LLAMA').
  • HIGHPublisher email 'you@example.com' is not affiliated with official vendor domain.
  • MEDIUMPackage published 5+ versions in rapid succession (< 24h) to mimic mature open source maintenance.
  • MEDIUMPackage registered 214 days ago within the 1-year AI tool proliferation window.
  • INFONo malicious install-time hooks or shell cradles detected in package AST.
  • INFOModerate community usage with 143 monthly downloads.

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 225d 17 71 ↑0.1/day
SUSPICIOUS
llama pypi 0.3.1 — you@example.com example.com 🪝
Package purpose: llama-index llms modelslab integration
Claimed homepage: https://pypi.org/project/llama-index-llms-modelslab/
Matched naming pattern: llama-llama-llms
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 225
Code-only verdict: BENIGN_COMMUNITY (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 17/month, 3/week (NEGLIGIBLE_OR_ZERO_USAGE)
Codebase size: 71 lines, 2.9 kB across 2 files (TINY_CODEBASE)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('LLAMA').
  • HIGHPublisher email 'you@example.com' is not affiliated with official vendor domain.
  • MEDIUMPackage published 5+ versions in rapid succession (< 24h) to mimic mature open source maintenance.
  • MEDIUMPackage registered 199 days ago within the 1-year AI tool proliferation window.
  • HIGHSOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in llama_index_llms_modelslab-0.3.1/llama_index/llms/modelslab/base.py:68
  • INFOModerate community usage with 186 monthly downloads.
Static code signals:
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in llama_index_llms_modelslab-0.3.1/llama_index/llms/modelslab/base.py:68

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 244d 167 24 ↑0.7/day
Possible Squatted Stub
langgraph pypi 1.0.0 — kdabhadk@gmail.com gmail.com
Package purpose: Renamed to langstage-core. This is a thin compatibility shim that re-exports langstage-core under the old import name.
Claimed homepage: https://pypi.org/project/langgraph-stream-parser/
Matched naming pattern: langgraph-langgraph-parser
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 244
Registry downloads: 167/month, 21/week (MODERATE_USAGE)
Codebase size: 24 lines, 1.2 kB across 1 file (EMPTY_STUB)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('LANGGRAPH').
  • HIGHPublisher email 'kdabhadk@gmail.com' is not affiliated with official vendor domain.
  • MEDIUMPackage registered 234 days ago within the 1-year AI tool proliferation window.
  • INFONo malicious install-time hooks or shell cradles detected in package AST.
  • MEDIUMPackage is an empty placeholder (24 LOC, 1 file(s), 1186 bytes).
  • INFOModerate community usage with 167 monthly downloads.

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 254d 154 67 ↑0.6/day
SUSPICIOUS
llama pypi 0.3.0 — roimkhan459@gmail.com gmail.com
Package purpose: llama-index tools mcp_discovery integration
Claimed homepage: https://pypi.org/project/llama-index-tools-mcp-discovery/
Matched naming pattern: llama-llama-tools
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 254
Code-only verdict: BENIGN_COMMUNITY (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 154/month, 18/week (MODERATE_USAGE)
Codebase size: 67 lines, 3.0 kB across 2 files (TINY_CODEBASE)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('LLAMA').
  • HIGHPublisher email 'roimkhan459@gmail.com' is not affiliated with official vendor domain.
  • MEDIUMPackage published 5+ versions in rapid succession (< 24h) to mimic mature open source maintenance.
  • MEDIUMPackage registered 243 days ago within the 1-year AI tool proliferation window.
  • INFONo malicious install-time hooks or shell cradles detected in package AST.
  • INFOModerate community usage with 154 monthly downloads.

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 263d 151 84 ↑0.6/day
SUSPICIOUS
llama pypi 0.2.0 Florian-BACHO f.bacho@outlook.fr outlook.fr
Package purpose: llama-index readers HuggingFace Datasets integration
Claimed homepage: https://pypi.org/project/llama-index-readers-datasets/
Matched naming pattern: llama-llama-readers
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 263
Code-only verdict: BENIGN_COMMUNITY (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 151/month, 17/week (MODERATE_USAGE)
Codebase size: 84 lines, 3.5 kB across 2 files (TINY_CODEBASE)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('LLAMA').
  • HIGHPublisher email 'f.bacho@outlook.fr' is not affiliated with official vendor domain.
  • MEDIUMPackage published 5+ versions in rapid succession (< 24h) to mimic mature open source maintenance.
  • MEDIUMPackage registered 255 days ago within the 1-year AI tool proliferation window.
  • INFONo malicious install-time hooks or shell cradles detected in package AST.
  • INFOModerate community usage with 151 monthly downloads.

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 292d 257 42 ↑0.9/day
SUSPICIOUS
llama pypi 0.3.1 — you@example.com example.com 🪝
Package purpose: llama-index llms aibadgr integration
Claimed homepage: https://pypi.org/project/llama-index-llms-aibadgr/
Matched naming pattern: llama-llama-llms
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 292
Code-only verdict: BENIGN_COMMUNITY (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 257/month, 19/week (MODERATE_USAGE)
Codebase size: 42 lines, 1.6 kB across 2 files (TINY_CODEBASE)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('LLAMA').
  • HIGHPublisher email 'you@example.com' is not affiliated with official vendor domain.
  • MEDIUMPackage published 5+ versions in rapid succession (< 24h) to mimic mature open source maintenance.
  • MEDIUMPackage registered 281 days ago within the 1-year AI tool proliferation window.
  • HIGHSOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in llama_index_llms_aibadgr-0.3.1/llama_index/llms/aibadgr/base.py:41
  • INFOModerate community usage with 257 monthly downloads.
Static code signals:
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in llama_index_llms_aibadgr-0.3.1/llama_index/llms/aibadgr/base.py:41

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 353d — — —
Unverified Signal
— npm latest PayAI Network — 🪝 🌐
Package purpose: A framework-agnostic implementation of the x402 payment protocol v2 for Solana clients (browsers) and servers
Static code signals:
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/dist/client/index.js:329 (+8 more occurrence(s) elsewhere)
  • SOURCE_CODE_ENCODED_PAYLOAD: 'Buffer.from(..., 'base64')' found in package/dist/index.js:444 (+4 more occurrence(s) elsewhere)
  • SUSPICIOUS_OBFUSCATION: 'Layered Base64 and Decompress Pipeline' found in package/dist/index.js:444 (+4 more occurrence(s) elsewhere)

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 361d 898 0 ↑2.5/day
Possible Squatted Stub
google pypi 0.0.1 — —
Package purpose: This is a placeholder package. Please install `google-cloud-aiplatform` instead.
Claimed homepage: https://pypi.org/project/google-cloud-vertexai/
Matched naming pattern: google-google-vertexai
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 361
Registry downloads: 898/month, 178/week (MODERATE_USAGE)
Codebase size: 0 lines, 0 Bytes across 11 files (EMPTY_STUB)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('GOOGLE').
  • HIGHPublisher email 'None' is not affiliated with official vendor domain.
  • MEDIUMPackage registered 353 days ago within the 1-year AI tool proliferation window.
  • INFONo malicious install-time hooks or shell cradles detected in package AST.
  • MEDIUMPackage is an empty placeholder (0 LOC, 11 file(s), 0 bytes).
  • INFOModerate community usage with 898 monthly downloads.

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 373d 263 72 ↑0.7/day
SUSPICIOUS
llama pypi 0.4.1 — yashp3020@gmail.com gmail.com
Package purpose: llama-index llms sarvam integration
Claimed homepage: https://pypi.org/project/llama-index-llms-sarvam/
Matched naming pattern: llama-llama-llms
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 373
Code-only verdict: BENIGN_COMMUNITY (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 263/month, 34/week (MODERATE_USAGE)
Codebase size: 72 lines, 2.8 kB across 2 files (TINY_CODEBASE)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('LLAMA').
  • HIGHPublisher email 'yashp3020@gmail.com' is not affiliated with official vendor domain.
  • MEDIUMPackage published 5+ versions in rapid succession (< 24h) to mimic mature open source maintenance.
  • MEDIUMPackage registered 362 days ago within the 1-year AI tool proliferation window.
  • INFONo malicious install-time hooks or shell cradles detected in package AST.
  • INFOModerate community usage with 263 monthly downloads.

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 394d 5 332 ↑0.0/day
SUSPICIOUS
coder pypi 1.3.1 — gargm0325@wrdsb.ca wrdsb.ca 🪝
Package purpose: A simple command line calculator
Claimed homepage: https://pypi.org/project/space-coder-cmd-calc/
Matched naming pattern: space-coder-cmd
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 394
Code-only verdict: Very Risky (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 5/month, 2/week (NEGLIGIBLE_OR_ZERO_USAGE)
Codebase size: 332 lines, 11.1 kB across 1 file (MODERATE_CODEBASE)
Automated signals flagged:
  • CRITICALMODULE_TOPLEVEL_EXECUTION: 'os.system' executed at module scope (runs on import) in space_coder_cmd_calc-1.3.1/src/cmd_calc/__main__.py:98
  • HIGHSOURCE_CODE_DYNAMIC_EXECUTION: 'Python Subprocess / OS Execution' found in space_coder_cmd_calc-1.3.1/src/cmd_calc/__main__.py:98
  • MEDIUMPackage executes code at install/import time, but no exfiltration destination, credential access, encoded payload, obfuscation, shell cradle, or persistence primitive corroborates a weaponized payload. Downgraded from MALICIOUS to SUSPICIOUS pending manual review.
Static code signals:
  • MODULE_TOPLEVEL_EXECUTION: 'os.system' executed at module scope (runs on import) in space_coder_cmd_calc-1.3.1/src/cmd_calc/__main__.py:98
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'Python Subprocess / OS Execution' found in space_coder_cmd_calc-1.3.1/src/cmd_calc/__main__.py:98

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 395d 7 254 ↑0.0/day
SUSPICIOUS
coder pypi 1.0.2 — gargm0325@wrdsb.ca wrdsb.ca 🪝
Package purpose: A simple calculator to keep track of you monopoly money
Claimed homepage: https://pypi.org/project/space-coder-monopoly-calc/
Matched naming pattern: space-coder-monopoly
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 395
Code-only verdict: Very Risky (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 7/month, 3/week (NEGLIGIBLE_OR_ZERO_USAGE)
Codebase size: 254 lines, 10.2 kB across 1 file (MODERATE_CODEBASE)
Automated signals flagged:
  • CRITICALMODULE_TOPLEVEL_EXECUTION: 'os.system' executed at module scope (runs on import) in space_coder_monopoly_calc-1.0.2/src/calc/__main__.py:13
  • HIGHSOURCE_CODE_DYNAMIC_EXECUTION: 'Python Subprocess / OS Execution' found in space_coder_monopoly_calc-1.0.2/src/calc/__main__.py:13
  • MEDIUMPackage executes code at install/import time, but no exfiltration destination, credential access, encoded payload, obfuscation, shell cradle, or persistence primitive corroborates a weaponized payload. Downgraded from MALICIOUS to SUSPICIOUS pending manual review.
Static code signals:
  • MODULE_TOPLEVEL_EXECUTION: 'os.system' executed at module scope (runs on import) in space_coder_monopoly_calc-1.0.2/src/calc/__main__.py:13
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'Python Subprocess / OS Execution' found in space_coder_monopoly_calc-1.0.2/src/calc/__main__.py:13

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 504d 216 131 ↑0.4/day
SUSPICIOUS
google npm 0.2.1 — —
Package purpose: A Model Context Protocol server for Google Programmable Search Engine (PSE)
Matched naming pattern: google-google-mcp
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 504
Code-only verdict: Possible Squatted Stub (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 216/month, 54/week (MODERATE_USAGE)
Codebase size: 131 lines, 5.2 kB across 1 file (TINY_CODEBASE)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('GOOGLE').
  • HIGHPublisher email 'None' is not affiliated with official vendor domain.
  • HIGHMISSING_SOURCE_REPOSITORY_URL
  • HIGHSOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/build/index.js:109
  • INFOModerate community usage with 216 monthly downloads.
Static code signals:
  • MISSING_SOURCE_REPOSITORY_URL
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/build/index.js:109

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 601d 166 136 ↑0.3/day
SUSPICIOUS
llama pypi 0.4.0 — you@example.com example.com 🪝 🗄️
Package purpose: llama-index retrievers kendra integration
Claimed homepage: https://pypi.org/project/llama-index-retrievers-kendra/
Matched naming pattern: llama-llama-retrievers
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 601
Code-only verdict: Unverified Signal (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 166/month, 18/week (MODERATE_USAGE)
Codebase size: 136 lines, 6.0 kB across 2 files (TINY_CODEBASE)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('LLAMA').
  • HIGHPublisher email 'you@example.com' is not affiliated with official vendor domain.
  • MEDIUMPackage published 5+ versions in rapid succession (< 24h) to mimic mature open source maintenance.
  • HIGHCREDENTIAL_PATH_HARVESTING: 'AWS Credentials (~/.aws/credentials)' found in llama_index_retrievers_kendra-0.4.0/llama_index/retrievers/kendra/base.py:21
  • INFOModerate community usage with 166 monthly downloads.
Static code signals:
  • CREDENTIAL_PATH_HARVESTING: 'AWS Credentials (~/.aws/credentials)' found in llama_index_retrievers_kendra-0.4.0/llama_index/retrievers/kendra/base.py:21

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 656d — — —
Unverified Signal
— npm latest Reown — 🪝 🌐
Package purpose: The full stack toolkit to build onchain app UX.
Static code signals:
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/dist/esm/src/utils/BitcoinApi.js:5
  • SOURCE_CODE_ENCODED_PAYLOAD: 'Buffer.from(..., 'base64')' found in package/dist/esm/src/connectors/UnisatConnector/index.js:64 (+3 more occurrence(s) elsewhere)
  • SUSPICIOUS_OBFUSCATION: 'Layered Base64 and Decompress Pipeline' found in package/dist/esm/src/connectors/UnisatConnector/index.js:64 (+2 more occurrence(s) elsewhere)

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 697d 648 109 ↑0.9/day
SUSPICIOUS
llama pypi 0.4.0 — ravi03071991@gmail.com gmail.com 🪝
Package purpose: llama-index postprocessor colpali-rerank integration
Claimed homepage: https://pypi.org/project/llama-index-postprocessor-colpali-rerank/
Matched naming pattern: llama-llama-postprocessor
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 697
Code-only verdict: BENIGN_COMMUNITY (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 648/month, 17/week (MODERATE_USAGE)
Codebase size: 109 lines, 4.7 kB across 2 files (TINY_CODEBASE)
Automated signals flagged:
  • HIGHClaims critical enterprise brand identity ('LLAMA').
  • HIGHPublisher email 'ravi03071991@gmail.com' is not affiliated with official vendor domain.
  • MEDIUMPackage published 5+ versions in rapid succession (< 24h) to mimic mature open source maintenance.
  • HIGHSOURCE_CODE_DYNAMIC_EXECUTION: 'eval()' found in llama_index_postprocessor_colpali_rerank-0.4.0/llama_index/postprocessor/colpali_rerank/base.py:51
  • INFOModerate community usage with 648 monthly downloads.
Static code signals:
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'eval()' found in llama_index_postprocessor_colpali_rerank-0.4.0/llama_index/postprocessor/colpali_rerank/base.py:51

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 1397d 8 183 ↑0.0/day
SUSPICIOUS
feature pypi 1.1.3 Amazon Web Services — 🪝
Package purpose: Amazon SageMaker FeatureStore PySpark Bindings
Claimed homepage: https://pypi.org/project/sagemaker-feature-store-pyspark-3.0/
Matched naming pattern: sagemaker-feature-store
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 1397
Code-only verdict: Very Risky (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 8/month, 0/week (NEGLIGIBLE_OR_ZERO_USAGE)
Codebase size: 183 lines, 10.5 kB across 4 files (MODERATE_CODEBASE)
Automated signals flagged:
  • INFOEstablished pre-AI legacy package registered 1371 days ago (predates modern AI hallucination waves).
  • CRITICALINSTALL_TIME_EXECUTION: 'subprocess.Popen' executed at top-level in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:61
  • HIGHSOURCE_CODE_ENV_VARS_ACCESS: Call to os.getenv() in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:11
  • HIGHSOURCE_CODE_ENV_VARS_ACCESS: Call to os.environ.get() in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:37
  • HIGHSOURCE_CODE_ENV_VARS_ACCESS: Access to os.environ in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:37
  • HIGHSOURCE_CODE_DYNAMIC_EXECUTION: 'Python Subprocess / OS Execution' found in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:61
  • HIGHSOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:11
  • HIGHINSTALL_TIME_EXECUTION: 'Python setup.py Custom Install Hook' found in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:34
Static code signals:
  • INSTALL_TIME_EXECUTION: 'subprocess.Popen' executed at top-level in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:61
  • SOURCE_CODE_ENV_VARS_ACCESS: Call to os.getenv() in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:11
  • SOURCE_CODE_ENV_VARS_ACCESS: Call to os.environ.get() in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:37
  • SOURCE_CODE_ENV_VARS_ACCESS: Access to os.environ in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:37
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'Python Subprocess / OS Execution' found in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:61
  • SOURCE_CODE_ENV_VARS_ACCESS: 'Environment Variable Access (process.env / os.environ)' found in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:11
  • INSTALL_TIME_EXECUTION: 'Python setup.py Custom Install Hook' found in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:34
  • SUPPLY_CHAIN_EXECUTION_HOOK: 'PyPI Custom Install Class Override' found in sagemaker_feature_store_pyspark_3_0-1.1.3/setup.py:34

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →
60 1542d — — —
Unverified Signal
— npm latest BitGo SDK Team — 🪝 🌐
Package purpose: BitGo SDK coin library for Sol
Static code signals:
  • SOURCE_CODE_ENCODED_PAYLOAD: 'Buffer.from(..., 'base64')' found in package/dist/cjs/src/lib/explainTransactionWasm.js:68 (+11 more occurrence(s) elsewhere)
  • SOURCE_CODE_NETWORK_CALL: 'fetch/axios/XMLHttpRequest' found in package/dist/cjs/src/lib/token2022Resolve.js:34 (+1 more occurrence(s) elsewhere)
  • SUSPICIOUS_OBFUSCATION: 'Layered Base64 and Decompress Pipeline' found in package/dist/cjs/src/lib/utils.js:152 (+1 more occurrence(s) elsewhere)

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on npm →
60 1733d 5 24 ↑0.0/day
SUSPICIOUS
base pypi 0.0.1 Ayan Khan programing.ninja0@gmail.com gmail.com 🪝
Package purpose: A demo project
Claimed homepage: https://pypi.org/project/base-calc/
Matched naming pattern: python-base-calc
Official vendor account: No — publisher domain doesn't match the brand it names
Days since publish: 1733
Code-only verdict: Very Risky (the score above also weighs naming and publisher-domain signals, not just the code itself)
Registry downloads: 5/month, 0/week (NEGLIGIBLE_OR_ZERO_USAGE)
Codebase size: 24 lines, 748 Bytes across 2 files (TINY_CODEBASE)
Automated signals flagged:
  • MEDIUMPackage has empty or near-zero description (potential reservation stub).
  • INFOEstablished pre-AI legacy package registered 1722 days ago (predates modern AI hallucination waves).
  • CRITICALMODULE_TOPLEVEL_EXECUTION: 'eval' executed at module scope (runs on import) in base_calc-0.0.1/base_calc/__init__.py:1
  • HIGHSOURCE_CODE_DYNAMIC_EXECUTION: 'eval()' found in base_calc-0.0.1/base_calc/__init__.py:1
  • MEDIUMPackage executes code at install/import time, but no exfiltration destination, credential access, encoded payload, obfuscation, shell cradle, or persistence primitive corroborates a weaponized payload. Downgraded from MALICIOUS to SUSPICIOUS pending manual review.
Static code signals:
  • MODULE_TOPLEVEL_EXECUTION: 'eval' executed at module scope (runs on import) in base_calc-0.0.1/base_calc/__init__.py:1
  • SOURCE_CODE_DYNAMIC_EXECUTION: 'eval()' found in base_calc-0.0.1/base_calc/__init__.py:1

Signals reflect what our detector observed, not a confirmed determination of wrongdoing.

View package on PyPI →

LINK COPIED TO CLIPBOARD