Back to skill

Security audit

Notion Export Desk

Security checks for vulnerabilities and agentic risk

Overview

The skill processes a user-selected Notion export locally and creates clearly disclosed reports without accessing accounts or transmitting content.

Select only the export you want processed and keep generated reports private because they contain copied page content. Retain the host sandbox. The bundled local workflow needs no Notion token or publisher account connection.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (5)

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: file_read, file_write, network, shell.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

The runtime implements outbound HTTP capabilities unrelated to indexing an existing local Notion export (L090–L115).

Medium
Category
Excessive Agency
Confidence
96% confidence
Finding

The manifest describes indexing existing Markdown/CSV exports, which does not require network access. The safe_request helper implements outbound requests with caller-supplied headers and bodies, including POST, PUT, PATCH, and DELETE (L092–L108). Although this file does not show the helper being invoked, the implemented network capability is outside the stated purpose.

Content

No source excerpt is available for this finding.

The HTTP helper can transmit request bodies and headers to remote hosts without user disclosure.

Medium
Category
Output Handling
Confidence
88% confidence
Finding

safe_request sends caller-supplied bodies and headers to the selected host, but the provided file contains no confirmation, user-facing network notice, or comment/docstring explaining this data transmission. HTTPS and destination validation protect the connection but do not inform users that their data may leave the workspace.

Content

No source excerpt is available for this finding.

Scope Creep

Low
Category
Excessive Agency
Confidence
70% confidence
Finding

Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.

Content

Scanner excerpt · LICENSE (reported line 7)May include surrounding context.

text
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

The report helper writes a file without any accompanying disclosure.

Low
Category
Output Handling
Confidence
90% confidence
Finding

The call to write_new persists content in the report directory. The helper includes no confirmation, user-facing log, docstring, or comment explaining that it creates a file.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.