Ai Cortex MCP Server
ai-cortex
npm
v0.17.0
Published by ai-creed — no publish provenance, so origin is unverified, but the source is public: the repository link below is self-declared yet readable, so you can inspect the code before adopting it.
Structural code intelligence and persistent memory for AI coding agents, served over MCP (Claude Code, Codex).
The grade answers one question — how safe is this server for you to adopt — so it is computed in two auditable stages. Nothing below is an opinion or an LLM's guess; every line is a real term the deterministic engine applied, and the same input always yields the same number.
1. Threat score — 100 − 0 = 100. What the published surface and source actually contain:
The deterministic scan raised no scored threat in the surface it inspected — the threat score stayed at 100. Capability observations and advisory notes are recorded but never lower it.
2. Client adoption risk — 100 − 7 = 93. Three small, subtract-only factors that reflect your risk in adopting it — a clean scan proves less on a powerful, unverified or barely-inspectable package, so the grade says so plainly:
| Points | Adoption-risk factor |
|---|---|
| −6 | capability blast radius (high) — client exposure if the model is manipulated |
| −1 | publisher verification (public source) — no provenance, but the source is public and inspectable |
Capability observations and info notes are shown under Findings but never scored.
Open any row's finding below for the file, line and evidence behind a deduction.
Findings 13
high Dynamic evaluation of a non-literal value (dist/web/graph/app3d.js)MTC-SRC-010
In the server's implementation (`dist/web/graph/app3d.js:75`): 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.
Evidence: ); let { Body } = new Function(code)(); return Body; } function generateCreateBodyFunctionBody(dimen
Fix: 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: server dist/web/graph/app3d.js
high Dynamic code execution in server code (dist/web/graph/app3d.js)MTC-SRC-001
In the server's implementation (`dist/web/graph/app3d.js:75`): 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.
Evidence: ); let { Body } = new Function(code)(); return Body; } function generateCreateBodyFunctionBody(dimen
Fix: 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: server dist/web/graph/app3d.js
high Shell/command execution in server code (dist/src/cli/graph.js)MTC-SRC-002
In the server's implementation (`dist/src/cli/graph.js:4`): 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.
Evidence: rt { spawn } from "node:child_process"; import { fileURLToPath } from "node:url"; import { loadRepoStores } from "../lib
Fix: 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: server dist/src/cli/graph.js
high Shell/command execution in server code (dist/src/lib/cache-store-migrate.js)MTC-SRC-002
In the server's implementation (`dist/src/lib/cache-store-migrate.js:4`): 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.
Evidence: ecFileSync } from "node:child_process"; import Database from "better-sqlite3"; export const SENTINEL_NAME = ".migration-
Fix: 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: server dist/src/lib/cache-store-migrate.js
high Shell/command execution in server code (dist/src/lib/cache-store.js)MTC-SRC-002
In the server's implementation (`dist/src/lib/cache-store.js:2`): 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.
Evidence: ort { exec } from "node:child_process"; import fs from "node:fs"; import os from "node:os"; import path from "node:path"
Fix: 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: server dist/src/lib/cache-store.js
high Shell/command execution in server code (dist/src/lib/diff-files.js)MTC-SRC-002
In the server's implementation (`dist/src/lib/diff-files.js:3`): 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.
Evidence: ort { exec } from "node:child_process"; import fs from "node:fs"; import path from "node:path"; import { promisify } fro
Fix: 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: server dist/src/lib/diff-files.js
high Shell/command execution in server code (dist/src/lib/indexable-files.js)MTC-SRC-002
In the server's implementation (`dist/src/lib/indexable-files.js:2`): 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.
Evidence: ort { exec } from "node:child_process"; import fs from "node:fs"; import path from "node:path"; import { promisify } fro
Fix: 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: server dist/src/lib/indexable-files.js
high Shell/command execution in server code (dist/src/lib/repo-identity.js)MTC-SRC-002
In the server's implementation (`dist/src/lib/repo-identity.js:2`): 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.
Evidence: ecFileSync } from "node:child_process"; import { createHash } from "node:crypto"; import fs from "node:fs"; import path
Fix: 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: server dist/src/lib/repo-identity.js
high Shell/command execution in server code (dist/src/lib/update-notifier.js)MTC-SRC-002
In the server's implementation (`dist/src/lib/update-notifier.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.
Evidence: rt { spawn } from "node:child_process"; const CACHE_TTL_MS = 24 * 60 * 60 * 1000; const FETCH_TIMEOUT_MS = 5_000; const
Fix: 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: server dist/src/lib/update-notifier.js
high Shell/command execution in server code (dist/src/server/graph-server.js)MTC-SRC-002
In the server's implementation (`dist/src/server/graph-server.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.
Evidence: rt { spawn } from "node:child_process"; import { loadRepoStores } from "../lib/graph/load.js"; import { buildGraph } fro
Fix: 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: server dist/src/server/graph-server.js
medium Tool "unlink_memories" can modify the filesystemMTC-CAP-002
Tool "unlink_memories" can write, overwrite or delete files (keyword "unlink" in tool name). Verify it is scoped to a safe directory.
Fix: Constrain file operations to an explicit, non-sensitive root; reject path traversal.
Location: tool unlink_memories
low Mutating tool "unlink_memories" declares no destructiveHintMTC-CAP-005
Tool "unlink_memories" can mutate/egress but declares no destructiveHint. Clients that don't default to spec-safe behavior may not prompt before running it.
Fix: Declare accurate annotations, and gate destructive tools on user confirmation regardless.
Location: tool unlink_memories
low Shell/command execution in packaging/dev tooling (dist/src/lib/library/__tests__/source-registry.test.js)MTC-SRC-002
In a packaging/dev/install script (shipped, but not the server runtime) (`dist/src/lib/library/__tests__/source-registry.test.js:2`): 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.
Evidence: ecFileSync } from "node:child_process"; import fs from "node:fs"; import os from "node:os"; import path from "node:path"
Fix: 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: server dist/src/lib/library/__tests__/source-registry.test.js
Tools 39
Each tool and what it can reach — statically extracted from the published source.
get_memoryingests untrusted inputrecall_memoryingests untrusted inputunlink_memorieswrites filesadd_evidenceno sensitive capabilityaudit_memoryno sensitive capabilityblast_radiusno sensitive capabilitycapture_sessionno sensitive capabilityconfirm_memoryno sensitive capabilitydeprecate_memoryno sensitive capabilityextract_sessionno sensitive capability
Show 29 more tools ↓
index_projectno sensitive capabilitylibrary_list_sourcesno sensitive capabilitylibrary_register_sourceno sensitive capabilitylibrary_reindexno sensitive capabilitylibrary_searchno sensitive capabilitylink_memoriesno sensitive capabilitylist_memoriesno sensitive capabilitylist_memories_pending_rewriteno sensitive capabilitymerge_memoriesno sensitive capabilitypin_memoryno sensitive capabilitypromote_to_globalno sensitive capabilitypurge_memoryno sensitive capabilityrebuild_indexno sensitive capabilityrecord_memoryno sensitive capabilityrehydrate_projectno sensitive capabilityrestore_memoryno sensitive capabilityreview_pending_capturesno sensitive capabilityrewrite_memoryno sensitive capabilitysearch_historyno sensitive capabilitysearch_memoriesno sensitive capabilitysuggest_filesno sensitive capabilitysuggest_files_deepno sensitive capabilitysuggest_files_semanticno sensitive capabilitysweep_agingno sensitive capabilitytrash_memoryno sensitive capabilityunpin_memoryno sensitive capabilityuntrash_memoryno sensitive capabilityupdate_memoryno sensitive capabilityupdate_scopeno sensitive capability
What this scan could not see
- Tools were statically extracted from the published source (39 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>".
Versions 1
Scan history per published version. The engine is deterministic — the same version always yields the same score, so a changed score means the package itself changed.
| Version | Score | Findings | Engine | Scanned |
|---|---|---|---|---|
v0.17.0 latest |
A 93/100 | 13 | 1.13.0 | 2026-09-07 |
Embed this score
Show this server's live Trust Score in your README, docs or website. The badge is served straight from the registry and updates automatically after every rescan — no API key needed. It links back to this page, so anyone who sees the grade can also read the findings behind it instead of taking a number on faith.
Verify this score yourself
The score above is reproducible: the same package version always yields the same result. Run it locally or over the free API — no account, no LLM, fully deterministic.
npx mcptrustchecker scan ai-cortex --onlineMore in AI, Memory & Reasoning
FDA device & vehicle recall risk for AI agents: recall history, MAUDE trend, risk score.
Open-source MCP server exposing Agent402.Tools' catalog — 500+ strong: 400+ self-hostable tools + 100 multi-tool skill packs (security-audit, trend-analysis, structured-scrape, decode-blob, forecasting-bake-off) for AI agents — browser, web search & answe
Zero-dependency MCP server that gives AI agents a self-updating project memory in AGENTS.md. Returns merge instructions instead of mutating state, so every change is a reviewable diff.
MCP Apps UI resources and server helpers for n8n
MCP server providing comprehensive access to BookStack knowledge management system
MCP server for tracking achievements with STAR methodology