Vault For Llm — the complete, unedited output of the deterministic mcptrustchecker engine v1.13.0, scanned . Every finding, capability tag and score component below is exactly what the engine produced — no AI, no post-processing.
{
"tool": {
"name": "mcptrustchecker",
"version": "1.13.0",
"methodologyVersion": "mcptrustchecker-1.13"
},
"target": {
"id": "vault-for-llm",
"source": {
"kind": "package",
"origin": "vault-for-llm"
},
"server": {
"name": "vault-for-llm"
}
},
"grade": "A",
"score": {
"score": 99,
"threatScore": 100,
"grade": "A",
"band": "A",
"categorySubtotals": {
"injection": 0,
"exfiltration": 0,
"permissions": 0,
"supply-chain": 0,
"network": 0,
"hygiene": 0
},
"vector": [
{
"kind": "client",
"term": "capability-exposure",
"level": "minimal",
"label": "capability blast radius (minimal) — client exposure if the model is manipulated",
"appliedPenalty": 0
},
{
"kind": "client",
"term": "verification-discount",
"level": "repo",
"label": "publisher verification (public source) — no provenance, but the source is public and inspectable",
"appliedPenalty": 1
},
{
"kind": "client",
"term": "coverage-honesty",
"level": "source",
"label": "inspection depth (source) — how much of the target the scan could see",
"appliedPenalty": 0
}
],
"gatesFired": [],
"methodologyVersion": "mcptrustchecker-1.13"
},
"capability": {
"level": "minimal",
"reasons": [],
"tags": []
},
"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-003",
"title": "Hardcoded egress to an external endpoint in packaging/dev tooling (scripts/cross_validate.py)",
"category": "exfiltration",
"severity": "low",
"confidence": "heuristic",
"description": "In a packaging/dev/install script (shipped, but not the server runtime) (`scripts/cross_validate.py:84`): A hardcoded outbound call to a fixed external host inside server code is a classic exfiltration/telemetry channel — especially paired with reads of local data. 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": "scripts/cross_validate.py"
},
"evidence": "try: resp = requests.post( \"https://ollama.com/v1/chat/completions\", headers={",
"owasp": "LLM02:2025 Sensitive Information Disclosure",
"data": {
"rule": "MTC-SRC-003",
"file": "scripts/cross_validate.py",
"line": 84,
"nonRuntime": true
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in packaging/dev tooling (scripts/convergence_check.py)",
"category": "permissions",
"severity": "low",
"confidence": "heuristic",
"description": "In a packaging/dev/install script (shipped, but not the server runtime) (`scripts/convergence_check.py:156`): 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": "scripts/convergence_check.py"
},
"evidence": "try: result = subprocess.run( [\"ollama\", \"run\", model, prompt], capture_output=True, t",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "scripts/convergence_check.py",
"line": 156,
"nonRuntime": true
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in packaging/dev tooling (scripts/daily_knowledge_sync.py)",
"category": "permissions",
"severity": "low",
"confidence": "heuristic",
"description": "In a packaging/dev/install script (shipped, but not the server runtime) (`scripts/daily_knowledge_sync.py:42`): 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": "scripts/daily_knowledge_sync.py"
},
"evidence": "try: result = subprocess.run( cmd, capture_output=True, text=True, timeout=timeout, cw",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "scripts/daily_knowledge_sync.py",
"line": 42,
"nonRuntime": true
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in packaging/dev tooling (scripts/history_privacy_scan.py)",
"category": "permissions",
"severity": "low",
"confidence": "heuristic",
"description": "In a packaging/dev/install script (shipped, but not the server runtime) (`scripts/history_privacy_scan.py:87`): 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": "scripts/history_privacy_scan.py"
},
"evidence": "rocess[str]: return subprocess.run( [\"git\", *args], cwd=repo, text=True, capture_out",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "scripts/history_privacy_scan.py",
"line": 87,
"nonRuntime": true
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in packaging/dev tooling (scripts/install_smoke_matrix.py)",
"category": "permissions",
"severity": "low",
"confidence": "heuristic",
"description": "In a packaging/dev/install script (shipped, but not the server runtime) (`scripts/install_smoke_matrix.py:47`): 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": "scripts/install_smoke_matrix.py"
},
"evidence": ") -> str: result = subprocess.run( command, cwd=cwd, env=env, input=input_text,",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "scripts/install_smoke_matrix.py",
"line": 47,
"nonRuntime": true
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in packaging/dev tooling (scripts/public_pr_gate.py)",
"category": "permissions",
"severity": "low",
"confidence": "heuristic",
"description": "In a packaging/dev/install script (shipped, but not the server runtime) (`scripts/public_pr_gate.py:260`): 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": "scripts/public_pr_gate.py"
},
"evidence": "clude).git\"] return subprocess.check_output(command, cwd=repo_root, text=True, errors=\"replace\") def render_text(r",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "scripts/public_pr_gate.py",
"line": 260,
"nonRuntime": true
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in packaging/dev tooling (scripts/readme_command_smoke.py)",
"category": "permissions",
"severity": "low",
"confidence": "heuristic",
"description": "In a packaging/dev/install script (shipped, but not the server runtime) (`scripts/readme_command_smoke.py:24`): 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": "scripts/readme_command_smoke.py"
},
"evidence": "]) -> str: result = subprocess.run( command, cwd=cwd, env=env, text=True, st",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "scripts/readme_command_smoke.py",
"line": 24,
"nonRuntime": true
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in packaging/dev tooling (scripts/search_qa_regression_gate.py)",
"category": "permissions",
"severity": "low",
"confidence": "heuristic",
"description": "In a packaging/dev/install script (shipped, but not the server runtime) (`scripts/search_qa_regression_gate.py:54`): 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": "scripts/search_qa_regression_gate.py"
},
"evidence": "output_path), ] subprocess.run(command, cwd=PROJECT_ROOT, check=True) return json.loads(output_path.read_tex",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "scripts/search_qa_regression_gate.py",
"line": 54,
"nonRuntime": true
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in packaging/dev tooling (scripts/watch_supabase_sync.py)",
"category": "permissions",
"severity": "low",
"confidence": "heuristic",
"description": "In a packaging/dev/install script (shipped, but not the server runtime) (`scripts/watch_supabase_sync.py:96`): 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": "scripts/watch_supabase_sync.py"
},
"evidence": ".env\")) completed = subprocess.run(command, text=True, capture_output=True) if completed.stdout: print(c",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "scripts/watch_supabase_sync.py",
"line": 96,
"nonRuntime": true
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in packaging/dev tooling (tests/test_agent_onboarding_benchmark.py)",
"category": "permissions",
"severity": "low",
"confidence": "heuristic",
"description": "In a packaging/dev/install script (shipped, but not the server runtime) (`tests/test_agent_onboarding_benchmark.py:139`): 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_agent_onboarding_benchmark.py"
},
"evidence": "mark.json\" result = subprocess.run( [ sys.executable, \"scripts/agent_onboarding_benc",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "tests/test_agent_onboarding_benchmark.py",
"line": 139,
"nonRuntime": true
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in packaging/dev tooling (tests/test_agent_onboarding_fixture_builder.py)",
"category": "permissions",
"severity": "low",
"confidence": "heuristic",
"description": "In a packaging/dev/install script (shipped, but not the server runtime) (`tests/test_agent_onboarding_fixture_builder.py:59`): 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_agent_onboarding_fixture_builder.py"
},
"evidence": "tmp_path): result = subprocess.run( [ sys.executable, \"scripts/build_agent_onboardin",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "tests/test_agent_onboarding_fixture_builder.py",
"line": 59,
"nonRuntime": true
}
},
{
"ruleId": "MTC-SRC-005",
"title": "Dynamic module load from a non-literal in packaging/dev tooling (tests/test_search_extended.py)",
"category": "permissions",
"severity": "low",
"confidence": "heuristic",
"description": "In a packaging/dev/install script (shipped, but not the server runtime) (`tests/test_search_extended.py:2637`): 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_search_extended.py"
},
"evidence": "s'\") return __import__(name) # Mock onnxruntime mock_ort = MagicMock() mock_session",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-005",
"file": "tests/test_search_extended.py",
"line": 2637,
"nonRuntime": true
}
},
{
"ruleId": "MTC-SUP-012",
"title": "Package has no license",
"category": "hygiene",
"severity": "info",
"confidence": "strong",
"description": "\"vault-for-llm\" has no declared license. This is a legal/reuse concern, not a security finding.",
"location": {
"kind": "package",
"name": "vault-for-llm"
}
}
],
"toxicFlows": [],
"capabilities": [],
"surfaceDigest": "b4992b4497353c43bc3fba118ef15c7e40e222835bf9e3c8a5428ef161608e96",
"stats": {
"tools": 0,
"prompts": 0,
"resources": 0,
"findingsBySeverity": {
"critical": 0,
"high": 0,
"medium": 0,
"low": 12,
"info": 1
}
}
}