registry / @ooples/token-optimizer-mcp / 7.0.1

@ooples/token-optimizer-mcp@7.0.1

Audit net MCP context avoided per AI coding agent, optimize context, and share a live local knowledge graph across 16 CLI clients

AI Security Review

scanned 14d ago · by lpm-firewall-ai

The source confirms conditional postinstall setup of package-owned agent integrations, which warrants a warning under the extension lifecycle policy. The inspected evidence does not establish malicious behavior or a publish-blocking attack.

Static reason

High-risk behavior combination matched malicious policy.; source matched previously finalized malicious package; routed for review

Trigger

npm postinstall during a global install outside CI.

Impact

Supported agent and shell integrations are configured automatically for qualifying installs.

Mechanism

The lifecycle script launches the package installer, which wires its hooks and records the installation.

AI rationale

Evidence

package.jsonscripts/postinstall.cjsscripts/install-cli.mjs

Decision evidence

public snapshot

AI called this Suspicious at 89.0% confidence as Dangerous Capability with medium false-positive risk.

Evidence for warning

package.json declares a postinstall script that invokes scripts/postinstall.cjs.
The postinstall setup runs only for global installs outside CI, then invokes the package installer.
The installer wires the package’s hooks and records the installation; this is guarded setup of package-owned integrations.
Raw transcript forwarding in the harvest feature requires an explicit environment flag.

Evidence against

The postinstall skips CI and local installs.
The inspected harvest code makes raw transcript forwarding opt-in.

Behavioral surface

Source

ChildProcessCryptoDynamicRequireEnvironmentVarsEvalFilesystemNetworkShell

Supply chain

HighEntropyStringsMinifiedTelemetryUrlStrings

ManifestNo manifest risk signals triggered.

scanned 1,050 file(s), 18.6 MB of source, external domains: 127.0.0.1, ai.google.dev, ampcode.com, api.anthropic.com, api.openai.com, chatgpt.com, cursor.com, developers.openai.com, docs.cline.bot, docs.continue.dev, docs.factory.ai, docs.github.com, docs.windsurf.com, evil.example, generativelanguage.googleapis.com, github.com, graphml.graphdrawing.org, kilo.ai, platform.claude.com, roocodeinc.github.io, www.gstatic.com, www.w3.org, zed.dev

Source & flagged code

18 flagged

package.jsonView file

Lines 33-73json

33 "integrations/**/*",

34 "plugin/**/*",

35 "scripts/**/*",

36 "!**/logs/**",

37 "!**/*.log",

38 "!**/.cache/**",

39 "!**/.token-optimizer/**",

40 "!**/*.db",

41 "!**/*.db-*",

42 "!**/*.sqlite",

43 "!**/*.sqlite-*",

44 "!**/*.sqlite3",

45 "!**/*.sqlite3-*",

46 "!**/*.tmp",

47 "!**/*.temp",

48 "!**/.token-optimizer-edit-*.lock",

49 "!**/.env",

50 "!**/.env.*"

51 ],

52 "scripts": {

53 "postinstall": "node scripts/postinstall.cjs",

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg · L53

Medium

Ambiguous Install Lifecycle Script

Install-time lifecycle script is not statically allowlisted and needs review.

package.jsonView on unpkg · L53

54 "checksums": "node scripts/checksums.mjs",

55 "screenshots": "node scripts/capture-screenshots.mjs",

56 "uninstall-hooks": "node scripts/uninstall.mjs",

57 "doctor": "node scripts/doctor.mjs",

58 "diagnostics": "node scripts/export-diagnostics.mjs",

59 "projects:discover": "node scripts/discover-project-graphs.mjs",

60 "build": "tsc && npm run copy:assets",

61 "copy:assets": "node scripts/copy-assets.mjs",

62 "start": "node dist/server/index.js",

63 "dev": "tsc --watch",

64 "clean": "rm -rf dist",

65 "prepublishOnly": "npm run clean && npm run build",

66 "test": "node --experimental-vm-modules node_modules/jest/bin/jest.js",

67 "test:watch": "node --experimental-vm-modules node_modules/jest/bin/jest.js --watch",

68 "test:ci": "node --experimental-vm-modules node_modules/jest/bin/jest.js --ci --coverage --maxWorkers=2",

69 "test:coverage": "node --experimental-vm-modules node_modules/jest/bin/jest.js --coverage",

70 "test:integration": "node --experimental-vm-modules node_modules/jest/bin/jest.js tests/integration",

71 "test:unit": "node --experimental-vm-modules node_modules/jest/bin/jest.js tests/unit",

72 "test:benchmark": "node --experimental-vm-modules node_modules/jest/bin/jest.js tests/benchmarks",

73 "dashboard": "node dist/server/web-server.js",

hooks-core/skeleton.mjsView file

Lines 9-49javascript

9 * can be strictly BETTER than the file:

10 *

11 * auth.ts -- 91 KB, structure + what we know

12 * function

hooks-core/harvest.mjsView file

Lines 4-44javascript

4 * Runs at Stop/PreCompact against a cheap model, in a separate process, so the

5 * session doing the work never pays for the harvest. Without an API key the

6 * whole thing is skipped and the graph keeps its structural layer, which is the

7 * majority of its machinery and costs nothing.

dist/compress/onnx.jsView file

Lines 41-81javascript

41 return hash % vocabSize;

42 });

43 };

44}

45/**

46 * Loads a model and returns an encoder over it.

47 *

48 * Throws if `onnxruntime-node` is absent or the model will not load, because

49 * this is called once at startup by a caller who explicitly asked for a model.

dist/tools/configuration/smart-env.jsView file

Lines 78-118javascript

78 await this.cacheResult(cacheKey, result);

79 const executionTime = Date.now() - startTime;

80 result.metadata.executionTime = executionTime;

81 this

hooks-core/harvest.mjs#virtual:normalized:round1View file

294// the prompt-file route exists to keep off a Windows command line: L295: // cmd.exe cannot carry a newline inside an argument, and the quoting L296: // doubles every `"`. These CLIs document the flag value as being\n // APPENDED to stdin, so the whole payload rides stdin and the argument\n // is one line of ASCII that any...

High

Command Output Exfiltration

Source combines command execution, command-output handling, and outbound requests; review data flow before blocking.

hooks-core/harvest.mjs#virtual:normalized:round1View on unpkg · L294

294Trigger-reachable command-output exfiltration chain: manifest.bin -> scripts/doctor.mjs -> hooks-core/doctor.mjs -> hooks-core/harvest.mjs L294: // the prompt-file route exists to keep off a Windows command line: L295: // cmd.exe cannot carry a newline inside an argument, and the quoting L296: // doubles every `"`. These CLIs document the flag value as being\n // APPENDED to stdin, so the whole payload rides stdin and the argument\n // is one line of ASCII that any...

High

Trigger Reachable Command Output Exfiltration

A manifest entrypoint or package-local install chain reaches command-output exfiltration behavior.

hooks-core/harvest.mjs#virtual:normalized:round1View on unpkg · L294

plugin/hooks/lib/restore.mjsView file

3contains invisible/control Unicode U+FEFF (zero width no-break space) <U+FEFF>/**

Critical

Trojan Source Unicode

Source contains bidi control or invisible Unicode characters associated with Trojan Source attacks.

plugin/hooks/lib/restore.mjsView on unpkg · L3

install-hooks.ps1View file

•path = install-hooks.ps1 kind = build_helper sizeBytes = 21337 magicHex = [redacted]

Medium

Ships Build Helper

Package ships non-JavaScript build or shell helper files.

install-hooks.ps1View on unpkg

dist/analysis/project-analyzer.jsView file

•stage = ast_semantic_analysis; reason = ast_alias_growth_limit_exceeded; limitedFiles = 23

High

Semantic Analysis Limited

A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.

dist/analysis/project-analyzer.jsView on unpkg

dist/server/daemon.jsView file

•matchType = normalized_sha256 matchedPackage = @ooples/token-optimizer-mcp@6.0.2 matchedPath = dist/server/daemon.js matchedIdentity = npm:[redacted]:6.0.2 similarity = 1.000 summary = normalized source hash matched finalized malicious source

High