Agentdb — 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": "agentdb",
"source": {
"kind": "package",
"origin": "agentdb"
},
"server": {
"name": "agentdb"
}
},
"grade": "A",
"score": {
"score": 92,
"threatScore": 99,
"grade": "A",
"band": "A",
"categorySubtotals": {
"injection": 0,
"exfiltration": 0,
"permissions": 0,
"supply-chain": 1.2,
"network": 0,
"hygiene": 0
},
"vector": [
{
"kind": "threat",
"ruleId": "MTC-SUP-010",
"category": "supply-chain",
"severity": "low",
"confidence": "heuristic",
"rawWeight": 3,
"confidenceMult": 0.4,
"diminishingFactor": 1,
"appliedPenalty": 1.2
},
{
"kind": "client",
"term": "capability-exposure",
"level": "high",
"label": "capability blast radius (high) — client exposure if the model is manipulated",
"appliedPenalty": 6
},
{
"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": "high",
"reasons": [
"can send data / act on an external service",
"can execute shell commands or code"
],
"tags": [
"code-exec",
"external-sink"
]
},
"coverage": {
"level": "source",
"inputs": {
"toolSurface": true,
"implementationSource": true,
"packageMetadata": true,
"liveTransport": false
},
"caveats": [
"Tools were statically extracted from the published source (35 recovered), not enumerated from a running server. Tool-poisoning, Unicode-smuggling, capability and toxic-flow analysis ran on this inferred surface, but a mis-parsed registration could be missed or mis-attributed, so tool-derived findings are capped below “confirmed”. To grade the real runtime surface, scan the running server: --command \"npx -y <package>\"."
]
},
"findings": [
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (dist/src/cli/agentdb-cli.js)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`dist/src/cli/agentdb-cli.js:1757`): 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": "dist/src/cli/agentdb-cli.js"
},
"evidence": "spawn } = await import('child_process'); const serverProcess = spawn('node', [mcpServerPath], { stdi",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "dist/src/cli/agentdb-cli.js",
"line": 1757,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (dist/src/cli/commands/install-embeddings.js)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`dist/src/cli/commands/install-embeddings.js:5`): 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": "dist/src/cli/commands/install-embeddings.js"
},
"evidence": "ort { spawnSync } from 'child_process'; // Color codes for beautiful output const colors = { reset: '\\x1b[0m', b",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "dist/src/cli/commands/install-embeddings.js",
"line": 5,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-002",
"title": "Shell/command execution in server code (dist/src/db-fallback.js)",
"category": "permissions",
"severity": "high",
"confidence": "strong",
"description": "In the server's implementation (`dist/src/db-fallback.js:377`): 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": "dist/src/db-fallback.js"
},
"evidence": "pper; } exec(sql) { return this.db.exec(sql); } save() { // Save",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-002",
"file": "dist/src/db-fallback.js",
"line": 377,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-003",
"title": "Hardcoded egress to an external endpoint (dist/src/controllers/EmbeddingService.js)",
"category": "exfiltration",
"severity": "medium",
"confidence": "strong",
"description": "In the server's implementation (`dist/src/controllers/EmbeddingService.js:214`): 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": "dist/src/controllers/EmbeddingService.js"
},
"evidence": "const response = await fetch('https://api.openai.com/v1/embeddings', { method: 'POST', headers:",
"owasp": "LLM02:2025 Sensitive Information Disclosure",
"data": {
"rule": "MTC-SRC-003",
"file": "dist/src/controllers/EmbeddingService.js",
"line": 214,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-003",
"title": "Hardcoded egress to an external endpoint (dist/src/services/LLMRouter.js)",
"category": "exfiltration",
"severity": "medium",
"confidence": "strong",
"description": "In the server's implementation (`dist/src/services/LLMRouter.js:383`): 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": "dist/src/services/LLMRouter.js"
},
"evidence": "const response = await fetch('https://openrouter.ai/api/v1/chat/completions', { method: 'POST',",
"owasp": "LLM02:2025 Sensitive Information Disclosure",
"data": {
"rule": "MTC-SRC-003",
"file": "dist/src/services/LLMRouter.js",
"line": 383,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-005",
"title": "Dynamic module load from a non-literal (dist/src/backends/ruvector/RuVectorBackend.js)",
"category": "permissions",
"severity": "medium",
"confidence": "heuristic",
"description": "In the server's implementation (`dist/src/backends/ruvector/RuVectorBackend.js:303`): 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": "dist/src/backends/ruvector/RuVectorBackend.js"
},
"evidence": "mmapModule = await import(moduleName); } catch { // mmap-io not available,",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-005",
"file": "dist/src/backends/ruvector/RuVectorBackend.js",
"line": 303,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-005",
"title": "Dynamic module load from a non-literal (dist/src/browser/HNSWIndex.js)",
"category": "permissions",
"severity": "medium",
"confidence": "heuristic",
"description": "In the server's implementation (`dist/src/browser/HNSWIndex.js:338`): 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": "dist/src/browser/HNSWIndex.js"
},
"evidence": "x from JSON */ import(json) { const data = JSON.parse(json); this.config = data.config;",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-005",
"file": "dist/src/browser/HNSWIndex.js",
"line": 338,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-005",
"title": "Dynamic module load from a non-literal (dist/src/cli/agentdb-cli.js)",
"category": "permissions",
"severity": "medium",
"confidence": "heuristic",
"description": "In the server's implementation (`dist/src/cli/agentdb-cli.js:1526`): 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": "dist/src/cli/agentdb-cli.js"
},
"evidence": "initScenario } = await import(runnerUrl); const subcommand = args[1]; if (!subcommand || subcom",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-005",
"file": "dist/src/cli/agentdb-cli.js",
"line": 1526,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-005",
"title": "Dynamic module load from a non-literal (dist/src/cli/commands/doctor.js)",
"category": "permissions",
"severity": "medium",
"confidence": "heuristic",
"description": "In the server's implementation (`dist/src/cli/commands/doctor.js:152`): 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": "dist/src/cli/commands/doctor.js"
},
"evidence": "try { require(mod.module); console.log(` ✅ ${mod.name} available`); passedChe",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-005",
"file": "dist/src/cli/commands/doctor.js",
"line": 152,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-005",
"title": "Dynamic module load from a non-literal (dist/src/cli/lib/config-manager.js)",
"category": "permissions",
"severity": "medium",
"confidence": "heuristic",
"description": "In the server's implementation (`dist/src/cli/lib/config-manager.js:498`): 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": "dist/src/cli/lib/config-manager.js"
},
"evidence": "SON string. */ import(json) { const config = JSON.parse(json); return this.validate(config);",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-005",
"file": "dist/src/cli/lib/config-manager.js",
"line": 498,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-005",
"title": "Dynamic module load from a non-literal (dist/src/cli/lib/report-store.js)",
"category": "permissions",
"severity": "medium",
"confidence": "heuristic",
"description": "In the server's implementation (`dist/src/cli/lib/report-store.js:392`): 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": "dist/src/cli/lib/report-store.js"
},
"evidence": "JSON. */ async import(json) { const results = JSON.parse(json); const ids = []; for (co",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-005",
"file": "dist/src/cli/lib/report-store.js",
"line": 392,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-005",
"title": "Dynamic module load from a non-literal (dist/src/cli/lib/simulation-registry.js)",
"category": "permissions",
"severity": "medium",
"confidence": "heuristic",
"description": "In the server's implementation (`dist/src/cli/lib/simulation-registry.js:99`): 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": "dist/src/cli/lib/simulation-registry.js"
},
"evidence": "const module = await import(filePath); const scenario = module.default || module;",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-005",
"file": "dist/src/cli/lib/simulation-registry.js",
"line": 99,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-005",
"title": "Dynamic module load from a non-literal (dist/src/cli/lib/simulation-runner.js)",
"category": "permissions",
"severity": "medium",
"confidence": "heuristic",
"description": "In the server's implementation (`dist/src/cli/lib/simulation-runner.js:107`): 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": "dist/src/cli/lib/simulation-runner.js"
},
"evidence": "const module = await import(fullPath); return module.default || module; }",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-005",
"file": "dist/src/cli/lib/simulation-runner.js",
"line": 107,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-005",
"title": "Dynamic module load from a non-literal (dist/src/controllers/WASMVectorSearch.js)",
"category": "permissions",
"severity": "medium",
"confidence": "heuristic",
"description": "In the server's implementation (`dist/src/controllers/WASMVectorSearch.js:129`): 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": "dist/src/controllers/WASMVectorSearch.js"
},
"evidence": "onst wasmModule = await import(normalizedPath); const ReasoningBankWasm = wasmModule.ReasoningBankWasm |",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-005",
"file": "dist/src/controllers/WASMVectorSearch.js",
"line": 129,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SRC-005",
"title": "Dynamic module load from a non-literal (dist/src/optimizations/ToolCache.js)",
"category": "permissions",
"severity": "medium",
"confidence": "heuristic",
"description": "In the server's implementation (`dist/src/optimizations/ToolCache.js:211`): 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": "dist/src/optimizations/ToolCache.js"
},
"evidence": "ersistence) */ import(entries) { const now = Date.now(); let imported = 0; for (const {",
"owasp": "LLM05:2025 Improper Output Handling",
"data": {
"rule": "MTC-SRC-005",
"file": "dist/src/optimizations/ToolCache.js",
"line": 211,
"nonRuntime": false
}
},
{
"ruleId": "MTC-SUP-010",
"title": "Package runs install-time scripts (postinstall)",
"category": "supply-chain",
"severity": "low",
"confidence": "heuristic",
"description": "\"agentdb\" executes postinstall script(s) at install time. An install hook runs at install time; most are routine build/setup, but review what it does before trusting it.",
"remediation": "Review the scripts; install with --ignore-scripts where possible and vet what they do.",
"location": {
"kind": "package",
"name": "agentdb"
},
"evidence": "node scripts/postinstall.cjs || true",
"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": [
{
"tool": "agentdb_init",
"tags": [],
"reasons": {}
},
{
"tool": "agentdb_insert",
"tags": [],
"reasons": {}
},
{
"tool": "agentdb_insert_batch",
"tags": [],
"reasons": {}
},
{
"tool": "agentdb_search",
"tags": [],
"reasons": {}
},
{
"tool": "agentdb_delete",
"tags": [],
"reasons": {}
},
{
"tool": "reflexion_store",
"tags": [],
"reasons": {}
},
{
"tool": "reflexion_retrieve",
"tags": [],
"reasons": {}
},
{
"tool": "skill_create",
"tags": [],
"reasons": {}
},
{
"tool": "skill_search",
"tags": [],
"reasons": {}
},
{
"tool": "causal_add_edge",
"tags": [],
"reasons": {}
},
{
"tool": "causal_query",
"tags": [],
"reasons": {}
},
{
"tool": "recall_with_certificate",
"tags": [],
"reasons": {}
},
{
"tool": "learner_discover",
"tags": [],
"reasons": {}
},
{
"tool": "db_stats",
"tags": [],
"reasons": {}
},
{
"tool": "learning_start_session",
"tags": [],
"reasons": {}
},
{
"tool": "learning_end_session",
"tags": [],
"reasons": {}
},
{
"tool": "learning_predict",
"tags": [],
"reasons": {}
},
{
"tool": "learning_feedback",
"tags": [],
"reasons": {}
},
{
"tool": "learning_train",
"tags": [],
"reasons": {}
},
{
"tool": "learning_metrics",
"tags": [],
"reasons": {}
},
{
"tool": "learning_transfer",
"tags": [],
"reasons": {}
},
{
"tool": "learning_explain",
"tags": [],
"reasons": {}
},
{
"tool": "experience_record",
"tags": [],
"reasons": {}
},
{
"tool": "reward_signal",
"tags": [],
"reasons": {}
},
{
"tool": "agentdb_stats",
"tags": [],
"reasons": {}
},
{
"tool": "agentdb_pattern_store",
"tags": [],
"reasons": {}
},
{
"tool": "agentdb_pattern_search",
"tags": [],
"reasons": {}
},
{
"tool": "agentdb_pattern_stats",
"tags": [],
"reasons": {}
},
{
"tool": "agentdb_clear_cache",
"tags": [],
"reasons": {}
},
{
"tool": "skill_create_batch",
"tags": [],
"reasons": {}
},
{
"tool": "reflexion_store_batch",
"tags": [],
"reasons": {}
},
{
"tool": "agentdb_pattern_store_batch",
"tags": [],
"reasons": {}
},
{
"tool": "causal_traverse",
"tags": [],
"reasons": {}
},
{
"tool": "agentdb_delete_batch",
"tags": [],
"reasons": {}
},
{
"tool": "consolidate_now",
"tags": [],
"reasons": {}
}
],
"surfaceDigest": "ef667065606f8dd2c6eaa9147a16f99941569d35c750b8c8ea7fbb5d7b58fac0",
"stats": {
"tools": 35,
"prompts": 0,
"resources": 0,
"findingsBySeverity": {
"critical": 0,
"high": 3,
"medium": 12,
"low": 1,
"info": 0
}
}
}