Agent Security Scanner — the complete, unedited output of the deterministic mcptrustchecker engine v1.5.0, scanned . Every finding, capability tag and score component below is exactly what the engine produced — no AI, no post-processing.
{
"tool": {
"name": "mcptrustchecker",
"version": "1.5.0",
"methodologyVersion": "mcptrustchecker-1.4"
},
"target": {
"id": "agent-security-scanner-mcp",
"source": {
"kind": "package",
"origin": "agent-security-scanner-mcp"
},
"server": {
"name": "agent-security-scanner-mcp"
}
},
"grade": "F",
"score": {
"score": 19,
"grade": "F",
"band": "F",
"gateCap": "C",
"categorySubtotals": {
"injection": 40.25,
"exfiltration": 22,
"permissions": 12.29,
"supply-chain": 6.3,
"network": 0,
"hygiene": 0
},
"vector": [
{
"ruleId": "MTC-SRC-007",
"category": "permissions",
"severity": "medium",
"confidence": "strong",
"rawWeight": 9,
"confidenceMult": 0.7,
"diminishingFactor": 1,
"appliedPenalty": 6.3
},
{
"ruleId": "MTC-SRC-007",
"category": "permissions",
"severity": "medium",
"confidence": "strong",
"rawWeight": 9,
"confidenceMult": 0.7,
"diminishingFactor": 0.5,
"appliedPenalty": 3.15
},
{
"ruleId": "MTC-SRC-007",
"category": "permissions",
"severity": "medium",
"confidence": "strong",
"rawWeight": 9,
"confidenceMult": 0.7,
"diminishingFactor": 0.25,
"appliedPenalty": 1.58
},
{
"ruleId": "MTC-SRC-007",
"category": "permissions",
"severity": "medium",
"confidence": "strong",
"rawWeight": 9,
"confidenceMult": 0.7,
"diminishingFactor": 0.1,
"appliedPenalty": 0.63
},
{
"ruleId": "MTC-SRC-007",
"category": "permissions",
"severity": "medium",
"confidence": "strong",
"rawWeight": 9,
"confidenceMult": 0.7,
"diminishingFactor": 0.1,
"appliedPenalty": 0.63
},
{
"ruleId": "MTC-SRC-008",
"category": "exfiltration",
"severity": "high",
"confidence": "confirmed",
"rawWeight": 22,
"confidenceMult": 1,
"diminishingFactor": 1,
"appliedPenalty": 22
},
{
"ruleId": "MTC-SRC-009",
"category": "injection",
"severity": "medium",
"confidence": "strong",
"rawWeight": 9,
"confidenceMult": 0.7,
"diminishingFactor": 1,
"appliedPenalty": 6.3
},
{
"ruleId": "MTC-SRC-009",
"category": "injection",
"severity": "medium",
"confidence": "strong",
"rawWeight": 9,
"confidenceMult": 0.7,
"diminishingFactor": 0.5,
"appliedPenalty": 3.15
},
{
"ruleId": "MTC-SRC-010",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 1,
"appliedPenalty": 15.4
},
{
"ruleId": "MTC-SRC-011",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 1,
"appliedPenalty": 15.4
},
{
"ruleId": "MTC-SUP-010",
"category": "supply-chain",
"severity": "medium",
"confidence": "strong",
"rawWeight": 9,
"confidenceMult": 0.7,
"diminishingFactor": 1,
"appliedPenalty": 6.3
}
],
"gatesFired": [
"1 confirmed high finding → grade capped at C"
],
"methodologyVersion": "mcptrustchecker-1.4"
},
"capability": {
"level": "high",
"reasons": [
"can execute shell commands or code"
],
"tags": [
"code-exec"
]
},
"coverage": {
"level": "source",
"inputs": {
"toolSurface": false,
"implementationSource": true,
"packageMetadata": true,
"liveTransport": false
},
"caveats": [
"No tools were enumerated, so prompt-injection, capability and toxic-flow analysis had no tool surface to inspect. To grade a package’s real runtime tools, scan the running server: --command \"npx -y <package>\"."
]
},
"findings": [
{
"ruleId": "MTC-SRC-008",
"title": "Hardcoded AWS access key id in server code (benchmarks/corpus/typescript_security.ts)",
"category": "exfiltration",
"severity": "high",
"confidence": "confirmed",
"description": "A live-looking AWS access key id is hardcoded in `benchmarks/corpus/typescript_security.ts:106`. Secrets in source ship to everyone who installs the package and are a direct credential leak.",
"remediation": "Remove the secret, rotate it, and load credentials from the environment or a secret store.",
"location": {
"kind": "server",
"name": "benchmarks/corpus/typescript_security.ts"
},
"evidence": "AWS access key id: AKIA…(redacted)",
"owasp": "LLM02:2025 Sensitive Information Disclosure",
"data": {
"secretType": "aws-access-key",
"file": "benchmarks/corpus/typescript_security.ts",
"line": 106
}
},
{
"ruleId": "MTC-SRC-010",
"title": "Dynamic evaluation of a non-literal value (pattern_matcher.py)",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`pattern_matcher.py:95`): Evaluating a runtime value as code (rather than a fixed literal) executes whatever reaches it — a direct RCE primitive, and almost never necessary in legitimate code. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "pattern_matcher.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-010",
"file": "pattern_matcher.py",
"line": 95
}
},
{
"ruleId": "MTC-SRC-011",
"title": "Assembled command execution and dynamic evaluation in the same server",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "The implementation builds shell commands out of runtime values AND evaluates runtime values as code. Each is a separate arbitrary-execution primitive; a server exposing both gives anything that reaches either one a direct path to running attacker-chosen code.",
"remediation": "Remove the dynamic eval, and pass command arguments as an argv array instead of building a shell string. If both are genuinely required, constrain and validate every value that can reach them.",
"location": {
"kind": "server",
"name": "implementation"
},
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rules": [
"MTC-SRC-009",
"MTC-SRC-010"
]
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (benchmarks/corpus/javascript_injection.js)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`benchmarks/corpus/javascript_injection.js:9`): Evaluating strings as code is the most direct RCE primitive; if any tool input reaches it, the server executes attacker-chosen code. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "benchmarks/corpus/javascript_injection.js"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "benchmarks/corpus/javascript_injection.js",
"line": 9
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (benchmarks/corpus/python_injection.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`benchmarks/corpus/python_injection.py:46`): Evaluating strings as code is the most direct RCE primitive; if any tool input reaches it, the server executes attacker-chosen code. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "benchmarks/corpus/python_injection.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "benchmarks/corpus/python_injection.py",
"line": 46
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (benchmarks/corpus/python_taint.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`benchmarks/corpus/python_taint.py:238`): Evaluating strings as code is the most direct RCE primitive; if any tool input reaches it, the server executes attacker-chosen code. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "benchmarks/corpus/python_taint.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "benchmarks/corpus/python_taint.py",
"line": 238
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (benchmarks/corpus/typescript_security.ts)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`benchmarks/corpus/typescript_security.ts:15`): Evaluating strings as code is the most direct RCE primitive; if any tool input reaches it, the server executes attacker-chosen code. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "benchmarks/corpus/typescript_security.ts"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "benchmarks/corpus/typescript_security.ts",
"line": 15
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (code-review-agent/dist/src/analyzer/intent.js)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`code-review-agent/dist/src/analyzer/intent.js:12`): Evaluating strings as code is the most direct RCE primitive; if any tool input reaches it, the server executes attacker-chosen code. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "code-review-agent/dist/src/analyzer/intent.js"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "code-review-agent/dist/src/analyzer/intent.js",
"line": 12
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (code-review-agent/dist/src/analyzer/semantic.js)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`code-review-agent/dist/src/analyzer/semantic.js:10`): Evaluating strings as code is the most direct RCE primitive; if any tool input reaches it, the server executes attacker-chosen code. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "code-review-agent/dist/src/analyzer/semantic.js"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "code-review-agent/dist/src/analyzer/semantic.js",
"line": 10
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (code-review-agent/src/analyzer/intent.ts)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`code-review-agent/src/analyzer/intent.ts:15`): Evaluating strings as code is the most direct RCE primitive; if any tool input reaches it, the server executes attacker-chosen code. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "code-review-agent/src/analyzer/intent.ts"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "code-review-agent/src/analyzer/intent.ts",
"line": 15
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (code-review-agent/src/analyzer/semantic.ts)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`code-review-agent/src/analyzer/semantic.ts:22`): Evaluating strings as code is the most direct RCE primitive; if any tool input reaches it, the server executes attacker-chosen code. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "code-review-agent/src/analyzer/semantic.ts"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "code-review-agent/src/analyzer/semantic.ts",
"line": 22
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (code-review-agent/tests/analyzer/engine.test.ts)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`code-review-agent/tests/analyzer/engine.test.ts:22`): Evaluating strings as code is the most direct RCE primitive; if any tool input reaches it, the server executes attacker-chosen code. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "code-review-agent/tests/analyzer/engine.test.ts"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "code-review-agent/tests/analyzer/engine.test.ts",
"line": 22
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (code-review-agent/tests/analyzer/semantic.test.ts)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`code-review-agent/tests/analyzer/semantic.test.ts:28`): Evaluating strings as code is the most direct RCE primitive; if any tool input reaches it, the server executes attacker-chosen code. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "code-review-agent/tests/analyzer/semantic.test.ts"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "code-review-agent/tests/analyzer/semantic.test.ts",
"line": 28
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (benchmarks/benchmark_runner.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`benchmarks/benchmark_runner.py:104`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "benchmarks/benchmark_runner.py"
},
"evidence": "subprocess.check_output(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "benchmarks/benchmark_runner.py",
"line": 104
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (benchmarks/corpus/javascript_injection.js)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`benchmarks/corpus/javascript_injection.js:25`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "benchmarks/corpus/javascript_injection.js"
},
"evidence": "child_process",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "benchmarks/corpus/javascript_injection.js",
"line": 25
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (benchmarks/corpus/python_injection.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`benchmarks/corpus/python_injection.py:27`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "benchmarks/corpus/python_injection.py"
},
"evidence": "subprocess.call(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "benchmarks/corpus/python_injection.py",
"line": 27
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (benchmarks/corpus/python_taint.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`benchmarks/corpus/python_taint.py:64`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "benchmarks/corpus/python_taint.py"
},
"evidence": "os.system(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "benchmarks/corpus/python_taint.py",
"line": 64
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (benchmarks/corpus/typescript_security.ts)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`benchmarks/corpus/typescript_security.ts:9`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "benchmarks/corpus/typescript_security.ts"
},
"evidence": "child_process",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "benchmarks/corpus/typescript_security.ts",
"line": 9
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (code-review-agent/dist/src/analyzer/intent.js)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`code-review-agent/dist/src/analyzer/intent.js:10`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "code-review-agent/dist/src/analyzer/intent.js"
},
"evidence": "subprocess.run(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "code-review-agent/dist/src/analyzer/intent.js",
"line": 10
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (code-review-agent/dist/src/analyzer/semantic.js)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`code-review-agent/dist/src/analyzer/semantic.js:9`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "code-review-agent/dist/src/analyzer/semantic.js"
},
"evidence": "subprocess.run(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "code-review-agent/dist/src/analyzer/semantic.js",
"line": 9
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (code-review-agent/dist/src/llm/claude-cli.js)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`code-review-agent/dist/src/llm/claude-cli.js:1`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "code-review-agent/dist/src/llm/claude-cli.js"
},
"evidence": "child_process",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "code-review-agent/dist/src/llm/claude-cli.js",
"line": 1
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (code-review-agent/src/analyzer/intent.ts)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`code-review-agent/src/analyzer/intent.ts:13`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "code-review-agent/src/analyzer/intent.ts"
},
"evidence": "subprocess.run(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "code-review-agent/src/analyzer/intent.ts",
"line": 13
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (code-review-agent/src/analyzer/semantic.ts)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`code-review-agent/src/analyzer/semantic.ts:21`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "code-review-agent/src/analyzer/semantic.ts"
},
"evidence": "subprocess.run(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "code-review-agent/src/analyzer/semantic.ts",
"line": 21
}
},
{
"ruleId": "MTC-SRC-009",
"title": "Untrusted input concatenated into a command sink (src/cli/demo.js)",
"category": "injection",
"severity": "medium",
"confidence": "strong",
"description": "In the server's implementation (`src/cli/demo.js:45`): A shell/process command assembled from concatenated or interpolated values is command injection when any part is attacker-influenced — the OWASP canonical RCE flow. Verify what reaches the interpolated value. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "src/cli/demo.js"
},
"evidence": "exec(\"ls \" +",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-009",
"file": "src/cli/demo.js",
"line": 45
}
},
{
"ruleId": "MTC-SRC-009",
"title": "Untrusted input concatenated into a command sink (src/fix-patterns.js)",
"category": "injection",
"severity": "medium",
"confidence": "strong",
"description": "In the server's implementation (`src/fix-patterns.js:68`): A shell/process command assembled from concatenated or interpolated values is command injection when any part is attacker-influenced — the OWASP canonical RCE flow. Verify what reaches the interpolated value. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "src/fix-patterns.js"
},
"evidence": "exec(\"cmd \" +",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-009",
"file": "src/fix-patterns.js",
"line": 68
}
},
{
"ruleId": "MTC-SRC-007",
"title": "Unsafe deserialization (benchmarks/corpus/python_injection.py)",
"category": "permissions",
"severity": "medium",
"confidence": "strong",
"description": "In the server's implementation (`benchmarks/corpus/python_injection.py:63`): Deserializing untrusted data with these APIs can execute arbitrary code (a well-known RCE gadget class). This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "benchmarks/corpus/python_injection.py"
},
"evidence": "pickle.loads(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-007",
"file": "benchmarks/corpus/python_injection.py",
"line": 63
}
},
{
"ruleId": "MTC-SRC-007",
"title": "Unsafe deserialization (benchmarks/corpus/python_taint.py)",
"category": "permissions",
"severity": "medium",
"confidence": "strong",
"description": "In the server's implementation (`benchmarks/corpus/python_taint.py:170`): Deserializing untrusted data with these APIs can execute arbitrary code (a well-known RCE gadget class). This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "benchmarks/corpus/python_taint.py"
},
"evidence": "pickle.loads(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-007",
"file": "benchmarks/corpus/python_taint.py",
"line": 170
}
},
{
"ruleId": "MTC-SRC-007",
"title": "Unsafe deserialization (src/cli/demo.js)",
"category": "permissions",
"severity": "medium",
"confidence": "strong",
"description": "In the server's implementation (`src/cli/demo.js:66`): Deserializing untrusted data with these APIs can execute arbitrary code (a well-known RCE gadget class). This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "src/cli/demo.js"
},
"evidence": "pickle.loads(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-007",
"file": "src/cli/demo.js",
"line": 66
}
},
{
"ruleId": "MTC-SRC-007",
"title": "Unsafe deserialization (src/fix-patterns.js)",
"category": "permissions",
"severity": "medium",
"confidence": "strong",
"description": "In the server's implementation (`src/fix-patterns.js:262`): Deserializing untrusted data with these APIs can execute arbitrary code (a well-known RCE gadget class). This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "src/fix-patterns.js"
},
"evidence": "node-serialize",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-007",
"file": "src/fix-patterns.js",
"line": 262
}
},
{
"ruleId": "MTC-SRC-007",
"title": "Unsafe deserialization (src/tools/scan-mcp.js)",
"category": "permissions",
"severity": "medium",
"confidence": "strong",
"description": "In the server's implementation (`src/tools/scan-mcp.js:318`): Deserializing untrusted data with these APIs can execute arbitrary code (a well-known RCE gadget class). This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.",
"remediation": "Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.",
"location": {
"kind": "server",
"name": "src/tools/scan-mcp.js"
},
"evidence": "yaml.load(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-007",
"file": "src/tools/scan-mcp.js",
"line": 318
}
},
{
"ruleId": "MTC-SUP-010",
"title": "Package runs install-time scripts (postinstall)",
"category": "supply-chain",
"severity": "medium",
"confidence": "strong",
"description": "\"agent-security-scanner-mcp\" executes postinstall script(s) at install time. The large majority of malicious packages abuse install hooks to run code before you ever import them.",
"remediation": "Review the scripts; install with --ignore-scripts where possible and vet what they do.",
"location": {
"kind": "package",
"name": "agent-security-scanner-mcp"
},
"owasp": "LLM03:2025 Supply Chain",
"references": [
"https://github.com/ossf/package-analysis",
"https://owasp.org/www-project-top-10-ci-cd-security-risks/"
],
"data": {
"scripts": [
"postinstall"
]
}
}
],
"toxicFlows": [],
"capabilities": [],
"surfaceDigest": "277cf92fc9fcebfa274c56284245fee11786d7f4cd1f857480c741bf0b009211",
"stats": {
"tools": 0,
"prompts": 0,
"resources": 0,
"findingsBySeverity": {
"critical": 0,
"high": 23,
"medium": 8,
"low": 0,
"info": 0
}
}
}