@sapiom/mcp@0.16.0
Local Sapiom developer MCP server (sapiom-dev) — the unmetered surface for building and operating on Sapiom; today it scaffolds, tests, deploys, and inspects orchestrations. Not the remote Sapiom capability MCP.
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThe package has no install-time execution or concrete payload, but its runtime design trusts remote model-facing instructions and exports complete tool arguments through default analytics. Those are unresolved data-exfiltration and agent-control risks.
Static reason
High-risk behavior combination matched malicious policy.; previous stored version diff introduced dangerous source
Trigger
Run the MCP entry point, then connect an MCP client or invoke any registered tool.
Impact
A compromised backend or analytics path could influence an attached agent or receive sensitive user-supplied tool arguments.
Mechanism
Remote instruction injection and default full-argument telemetry.
AI rationale
The package has no install-time execution or concrete payload, but its runtime design trusts remote model-facing instructions and exports complete tool arguments through default analytics. Those are unresolved data-exfiltration and agent-control risks.
Evidence
dist/index.jsdist/instructions-fetch.jsdist/analytics.jsdist/register-tool.js~/.sapiom/credentials.json
Network endpoints2
api.sapiom.ai/v1/mcp/instructionsapi.sapiom.dev/v1/mcp/instructions
Decision evidence
public snapshotAI called this Suspicious at 89.0% confidence as Dangerous Capability with low false-positive risk.
Evidence for warning
- On startup, it fetches remotely controlled text and presents it as MCP instructions.
- Analytics starts by default with a cached API key.
- Each tool call forwards the complete argument object to analytics, which can include sensitive project data or secrets.
Evidence against
- There are no npm install lifecycle hooks.
- Browser and credential writes occur through the explicit authentication tool.
- The OAuth callback verifies a random state value on localhost.
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsUrlStrings
Source & flagged code
7 flaggeddist/auth.jsView file
1import * as http from "node:http";
L2: import * as crypto from "node:crypto";
L3: import { spawn } from "node:child_process";
L4: import { URL, URLSearchParams } from "node:url";
L5: /**
...
L9: * - NEVER a visible console. The previous `execSync('start "" "<url>"')`
L10: * spawned cmd.exe, which pops a console window on the user's screen when
L11: * the caller isn't attached to one (an MCP server child of a GUI app
...
L16: * a cmd.exe command line treats as a command separator unless quoted
L17: * exactly right. The PowerShell command is passed BASE64-ENCODED
L18: * (-EncodedCommand), so no shell ever tokenizes the URL; the
...
L21: *
Critical
Reverse Shell
Source matches reverse-shell style process and socket wiring.
dist/auth.jsView on unpkg · L11Trigger-reachable chain: manifest.exports -> dist/auth-api.d.ts -> dist/auth.js
L1: import * as http from "node:http";
L2: import * as crypto from "node:crypto";
L3: import { spawn } from "node:child_process";
L4: import { URL, URLSearchParams } from "node:url";
L5: /**
...
L9: * - NEVER a visible console. The previous `execSync('start "" "<url>"')`
L10: * spawned cmd.exe, which pops a console window on the user's screen when
L11: * the caller isn't attached to one (an MCP server child of a GUI app
...
L16: * a cmd.exe command line treats as a command separator unless quoted
L17: * exactly right. The PowerShell command is passed BASE64-ENCODED
L18: * (-EncodedCommand), so no shell ever tokenizes the URL; the
...
L21: *
Critical
Trigger Reachable Dangerous Capability
A package entrypoint or install-time lifecycle script reaches a source file with blocking dangerous behavior.
Findings
3 Critical4 High3 Medium4 Low
CriticalReverse Shelldist/auth.js
CriticalTrigger Reachable Dangerous Capabilitydist/auth.js
CriticalPrevious Version Dangerous Deltadist/auth.js
HighChild Processdist/auth.js
HighShelldist/auth.js
HighSandbox Evasion Gated Capabilitydist/auth.js
HighSemantic Analysis Limiteddist/tools/app-manage.js#virtual:normalized:round1
MediumNetwork
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
Affected versions and remediation
This report applies to @sapiom/mcp@0.16.0.
See all versions of @sapiom/mcp for other recorded verdicts.
- Review the evidence and your use of @sapiom/mcp@0.16.0 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.