Cortexgraph MCP Server
cortexgraph
PyPI
v1.2.1
Published by an unidentified publisher — no publish provenance and no public repository, so the publisher could not be verified and the source cannot be independently located.
CortexGraph: Temporal memory management for AI assistants with human-like dynamics
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 − 1.2 = 99. What the published surface and source actually contain:
| Points | What was found | Category |
|---|---|---|
| −1.2 | Hardcoded JSON Web Token in test/example/packaging MTC-SRC-008 | exfiltration |
2. Client adoption risk — 99 − 8 = 91. 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 |
| −2 | publisher verification (unlinked) — no provenance/repo link, but the shipped source was fully read |
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 4
high Dynamic evaluation of a non-literal value (site/assets/javascripts/lunr/wordcut.js)MTC-SRC-010
In the server's implementation (`site/assets/javascripts/lunr/wordcut.js:5977`): 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: ry { var render = new Function(settings.variable || 'obj', '_', source); } catch (e) { e.source = source
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 site/assets/javascripts/lunr/wordcut.js
high Dynamic code execution in server code (site/assets/javascripts/lunr/wordcut.js)MTC-SRC-001
In the server's implementation (`site/assets/javascripts/lunr/wordcut.js:5977`): 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: ry { var render = new Function(settings.variable || 'obj', '_', source); } catch (e) { e.source = source
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 site/assets/javascripts/lunr/wordcut.js
high Shell/command execution in server code (src/cortexgraph/agents/beads_integration.py)MTC-SRC-002
In the server's implementation (`src/cortexgraph/agents/beads_integration.py:68`): 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: try: result = subprocess.run( cmd, capture_output=True, text=True,
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 src/cortexgraph/agents/beads_integration.py
low Hardcoded JSON Web Token in test/example/packaging (tests/test_security_secrets.py)MTC-SRC-008
A hardcoded JSON Web Token (an embedded JSON Web Token (often an example or expired token)) appears in `tests/test_security_secrets.py:377`. Verify whether this is a real credential; if so, remove and rotate it.
Evidence: JSON Web Token: eyJh…(redacted)
Fix: Remove the secret, rotate it, and load credentials from the environment or a secret store.
Location: server tests/test_security_secrets.py
Tools 13
Each tool and what it can reach — statically extracted from the published source.
auto_recall_process_messageno sensitive capabilitybackfill_embeddingsno sensitive capabilitycluster_memoriesno sensitive capabilityconsolidate_memoriesno sensitive capabilitycreate_relationno sensitive capabilitygcno sensitive capabilityget_performance_metricsno sensitive capabilityobserve_memory_usageno sensitive capabilityopen_memoriesno sensitive capabilitypromote_memoryno sensitive capability
Show 3 more tools ↓
read_graphno sensitive capabilityreset_performance_metricsno sensitive capabilitytouch_memoryno sensitive capability
What this scan could not see
- Tools were statically extracted from the published source (13 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 |
|---|---|---|---|---|
v1.2.1 latest |
A 91/100 | 4 | 1.13.0 | 2026-08-25 |
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 cortexgraph --online --registry pypiMore in Learning & Documentation
stdio compatibility bridge for the hosted AgentMail MCP server
MCP server for Alfa documentation retrieval
MCP server providing Alpine.js documentation and reference
Search the official Astro framework documentation.
MCP server providing ATAG 2.0 (Authoring Tool Accessibility Guidelines) guidance for authoring tool developers
Aurais Research Reader as an MCP server — paste a paper or article, get a structured read with verbatim-verified quotes + a signed Vorion proof chain (via @vorionsys/aurais-core).