Vmlx — 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": "vmlx",
"source": {
"kind": "package",
"origin": "vmlx"
},
"server": {
"name": "vmlx"
}
},
"grade": "F",
"score": {
"score": 48,
"grade": "F",
"band": "F",
"categorySubtotals": {
"injection": 50,
"exfiltration": 0,
"permissions": 0,
"supply-chain": 2.1,
"network": 0,
"hygiene": 0
},
"vector": [
{
"ruleId": "MTC-SRC-004",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 1,
"appliedPenalty": 15.4
},
{
"ruleId": "MTC-SRC-010",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 1,
"appliedPenalty": 15.4
},
{
"ruleId": "MTC-SRC-010",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 0.5,
"appliedPenalty": 7.7
},
{
"ruleId": "MTC-SRC-010",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 0.25,
"appliedPenalty": 3.85
},
{
"ruleId": "MTC-SRC-010",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 0.1,
"appliedPenalty": 1.54
},
{
"ruleId": "MTC-SRC-010",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 0.1,
"appliedPenalty": 1.54
},
{
"ruleId": "MTC-SRC-010",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 0.1,
"appliedPenalty": 1.54
},
{
"ruleId": "MTC-SRC-010",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 0.1,
"appliedPenalty": 1.54
},
{
"ruleId": "MTC-SRC-010",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 0.1,
"appliedPenalty": 1.54
},
{
"ruleId": "MTC-SRC-010",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 0.1,
"appliedPenalty": 1.54
},
{
"ruleId": "MTC-SRC-010",
"category": "injection",
"severity": "high",
"confidence": "strong",
"rawWeight": 22,
"confidenceMult": 0.7,
"diminishingFactor": 0.1,
"appliedPenalty": 1.54
},
{
"ruleId": "MTC-SUP-011",
"category": "supply-chain",
"severity": "low",
"confidence": "strong",
"rawWeight": 3,
"confidenceMult": 0.7,
"diminishingFactor": 1,
"appliedPenalty": 2.1
}
],
"gatesFired": [],
"methodologyVersion": "mcptrustchecker-1.4"
},
"capability": {
"level": "high",
"reasons": [
"reads sensitive or local data",
"has a read → egress path (a data-exfiltration surface)",
"can execute shell commands or code"
],
"tags": [
"sensitive-source",
"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-006",
"title": "Reads a sensitive credential path or dumps the environment (tests/test_release_gate_python_app.py)",
"category": "exfiltration",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_release_gate_python_app.py:850`): Reading private keys / cloud credentials, or serializing the whole environment, is a sensitive-data source that becomes exfiltration when combined with any egress. 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": "tests/test_release_gate_python_app.py"
},
"evidence": "dict(os.environ)",
"owasp": "LLM02:2025 Sensitive Information Disclosure",
"data": {
"rule": "MTC-SRC-006",
"file": "tests/test_release_gate_python_app.py",
"line": 850
}
},
{
"ruleId": "MTC-SRC-006",
"title": "Reads a sensitive credential path or dumps the environment (vmlx_engine/native_mtp_examples/server_smoke.py)",
"category": "exfiltration",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`vmlx_engine/native_mtp_examples/server_smoke.py:102`): Reading private keys / cloud credentials, or serializing the whole environment, is a sensitive-data source that becomes exfiltration when combined with any egress. 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": "vmlx_engine/native_mtp_examples/server_smoke.py"
},
"evidence": "dict(os.environ)",
"owasp": "LLM02:2025 Sensitive Information Disclosure",
"data": {
"rule": "MTC-SRC-006",
"file": "vmlx_engine/native_mtp_examples/server_smoke.py",
"line": 102
}
},
{
"ruleId": "MTC-SRC-004",
"title": "Obfuscated / encoded payload in server code (tests/test_image_api.py)",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_image_api.py:470`): Decoding an encoded blob and executing it is how malicious logic is hidden from human review of the source. 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": "tests/test_image_api.py"
},
"evidence": "\\x00\\x00\\x00\\x01\\x08\\x06\\x00\\x00\\x00\\x1f\\x15\\xc4",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-004",
"file": "tests/test_image_api.py",
"line": 470
}
},
{
"ruleId": "MTC-SRC-010",
"title": "Dynamic evaluation of a non-literal value (vmlx_engine/block_disk_store.py)",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`vmlx_engine/block_disk_store.py:628`): 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": "vmlx_engine/block_disk_store.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-010",
"file": "vmlx_engine/block_disk_store.py",
"line": 628
}
},
{
"ruleId": "MTC-SRC-010",
"title": "Dynamic evaluation of a non-literal value (vmlx_engine/disk_cache.py)",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`vmlx_engine/disk_cache.py:947`): 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": "vmlx_engine/disk_cache.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-010",
"file": "vmlx_engine/disk_cache.py",
"line": 947
}
},
{
"ruleId": "MTC-SRC-010",
"title": "Dynamic evaluation of a non-literal value (vmlx_engine/loaders/load_zaya.py)",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`vmlx_engine/loaders/load_zaya.py:43`): 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": "vmlx_engine/loaders/load_zaya.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-010",
"file": "vmlx_engine/loaders/load_zaya.py",
"line": 43
}
},
{
"ruleId": "MTC-SRC-010",
"title": "Dynamic evaluation of a non-literal value (vmlx_engine/logprobs.py)",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`vmlx_engine/logprobs.py:40`): 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": "vmlx_engine/logprobs.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-010",
"file": "vmlx_engine/logprobs.py",
"line": 40
}
},
{
"ruleId": "MTC-SRC-010",
"title": "Dynamic evaluation of a non-literal value (vmlx_engine/memory_cache.py)",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`vmlx_engine/memory_cache.py:575`): 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": "vmlx_engine/memory_cache.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-010",
"file": "vmlx_engine/memory_cache.py",
"line": 575
}
},
{
"ruleId": "MTC-SRC-010",
"title": "Dynamic evaluation of a non-literal value (vmlx_engine/mllm_batch_generator.py)",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`vmlx_engine/mllm_batch_generator.py:2536`): 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": "vmlx_engine/mllm_batch_generator.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-010",
"file": "vmlx_engine/mllm_batch_generator.py",
"line": 2536
}
},
{
"ruleId": "MTC-SRC-010",
"title": "Dynamic evaluation of a non-literal value (vmlx_engine/mllm_scheduler.py)",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`vmlx_engine/mllm_scheduler.py:1538`): 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": "vmlx_engine/mllm_scheduler.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-010",
"file": "vmlx_engine/mllm_scheduler.py",
"line": 1538
}
},
{
"ruleId": "MTC-SRC-010",
"title": "Dynamic evaluation of a non-literal value (vmlx_engine/model_runner.py)",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`vmlx_engine/model_runner.py:335`): 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": "vmlx_engine/model_runner.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-010",
"file": "vmlx_engine/model_runner.py",
"line": 335
}
},
{
"ruleId": "MTC-SRC-010",
"title": "Dynamic evaluation of a non-literal value (vmlx_engine/models/gemma4_unified/gemma4_unified.py)",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`vmlx_engine/models/gemma4_unified/gemma4_unified.py:298`): 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": "vmlx_engine/models/gemma4_unified/gemma4_unified.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-010",
"file": "vmlx_engine/models/gemma4_unified/gemma4_unified.py",
"line": 298
}
},
{
"ruleId": "MTC-SRC-010",
"title": "Dynamic evaluation of a non-literal value (vmlx_engine/models/minimax_m3/eagle3_decode.py)",
"category": "injection",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`vmlx_engine/models/minimax_m3/eagle3_decode.py:61`): 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": "vmlx_engine/models/minimax_m3/eagle3_decode.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-010",
"file": "vmlx_engine/models/minimax_m3/eagle3_decode.py",
"line": 61
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (tests/test_batching.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_batching.py:262`): 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": "tests/test_batching.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "tests/test_batching.py",
"line": 262
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (tests/test_block_disk_bfloat16.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_block_disk_bfloat16.py:41`): 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": "tests/test_block_disk_bfloat16.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "tests/test_block_disk_bfloat16.py",
"line": 41
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (tests/test_cache_isolation.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_cache_isolation.py:247`): 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": "tests/test_cache_isolation.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "tests/test_cache_isolation.py",
"line": 247
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (tests/test_continuous_batching_extend.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_continuous_batching_extend.py:43`): 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": "tests/test_continuous_batching_extend.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "tests/test_continuous_batching_extend.py",
"line": 43
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (tests/test_dsv4_batch_generator_speed.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_dsv4_batch_generator_speed.py: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": "tests/test_dsv4_batch_generator_speed.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "tests/test_dsv4_batch_generator_speed.py",
"line": 28
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (tests/test_dsv4_paged_cache.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_dsv4_paged_cache.py:1139`): 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": "tests/test_dsv4_paged_cache.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "tests/test_dsv4_paged_cache.py",
"line": 1139
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (tests/test_hybrid_live_tq_kv.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_hybrid_live_tq_kv.py:685`): 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": "tests/test_hybrid_live_tq_kv.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "tests/test_hybrid_live_tq_kv.py",
"line": 685
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (tests/test_impl_campaign_20260710.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_impl_campaign_20260710.py:234`): 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": "tests/test_impl_campaign_20260710.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "tests/test_impl_campaign_20260710.py",
"line": 234
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (tests/test_jang_affine_storage.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_jang_affine_storage.py:73`): 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": "tests/test_jang_affine_storage.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "tests/test_jang_affine_storage.py",
"line": 73
}
},
{
"ruleId": "MTC-SRC-001",
"title": "Dynamic code execution in server code (tests/test_jang_loader.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_jang_loader.py:333`): 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": "tests/test_jang_loader.py"
},
"evidence": "eval(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-001",
"file": "tests/test_jang_loader.py",
"line": 333
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (tests/test_api_surface_contract.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_api_surface_contract.py:115`): 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": "tests/test_api_surface_contract.py"
},
"evidence": "child_process",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "tests/test_api_surface_contract.py",
"line": 115
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (tests/test_cache_bypass.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_cache_bypass.py:1080`): 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": "tests/test_cache_bypass.py"
},
"evidence": "subprocess.run(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "tests/test_cache_bypass.py",
"line": 1080
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (tests/test_current_regression_suite.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_current_regression_suite.py:625`): 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": "tests/test_current_regression_suite.py"
},
"evidence": "subprocess.Popen(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "tests/test_current_regression_suite.py",
"line": 625
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (tests/test_e2e_live.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_e2e_live.py:244`): 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": "tests/test_e2e_live.py"
},
"evidence": "subprocess.Popen(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "tests/test_e2e_live.py",
"line": 244
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (tests/test_full_release_objective_checklist.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_full_release_objective_checklist.py:51`): 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": "tests/test_full_release_objective_checklist.py"
},
"evidence": "child_process",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "tests/test_full_release_objective_checklist.py",
"line": 51
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (tests/test_installed_app_runtime_parity_audit.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_installed_app_runtime_parity_audit.py:28`): 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": "tests/test_installed_app_runtime_parity_audit.py"
},
"evidence": "child_process",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "tests/test_installed_app_runtime_parity_audit.py",
"line": 28
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (tests/test_native_mtp_autodetect.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_native_mtp_autodetect.py:227`): 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": "tests/test_native_mtp_autodetect.py"
},
"evidence": "subprocess.run(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "tests/test_native_mtp_autodetect.py",
"line": 227
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (tests/test_panel_cli_flag_contract.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_panel_cli_flag_contract.py:393`): 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": "tests/test_panel_cli_flag_contract.py"
},
"evidence": "spawn(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "tests/test_panel_cli_flag_contract.py",
"line": 393
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (tests/test_release_gate_python_app.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_release_gate_python_app.py:852`): 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": "tests/test_release_gate_python_app.py"
},
"evidence": "subprocess.run(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "tests/test_release_gate_python_app.py",
"line": 852
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (tests/test_tq_e2e_live.py)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`tests/test_tq_e2e_live.py:348`): 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": "tests/test_tq_e2e_live.py"
},
"evidence": "subprocess.Popen(",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "tests/test_tq_e2e_live.py",
"line": 348
}
},
{
"ruleId": "MTC-SRC-005",
"title": "Dynamic module load from a non-literal (tests/test_tq_disk_cache.py)",
"category": "permissions",
"severity": "medium",
"confidence": "heuristic",
"description": "In the server's implementation (`tests/test_tq_disk_cache.py:397`): Loading a module chosen at runtime (from a variable) can pull in and run attacker-influenced code paths. 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": "tests/test_tq_disk_cache.py"
},
"evidence": "__import__(\n ",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-005",
"file": "tests/test_tq_disk_cache.py",
"line": 397
}
},
{
"ruleId": "MTC-SUP-011",
"title": "Package has no source repository",
"category": "supply-chain",
"severity": "low",
"confidence": "strong",
"description": "\"vmlx\" declares no repository URL, so its published artifact cannot be compared against reviewable source.",
"remediation": "Prefer packages that link to public, reviewable source.",
"location": {
"kind": "package",
"name": "vmlx"
}
}
],
"toxicFlows": [],
"capabilities": [],
"surfaceDigest": "d221491cfdbba02bf3e186db24cf517d245ff29fddcfc8de94602c5fee2f5530",
"stats": {
"tools": 0,
"prompts": 0,
"resources": 0,
"findingsBySeverity": {
"critical": 0,
"high": 33,
"medium": 1,
"low": 1,
"info": 0
}
}
}