Back to skill

Security audit

Retrospective Analysis

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward retrospective-analysis helper with disclosed, user-directed JSON save/load features and no evidence of hidden execution, network access, credential use, or destructive behavior.

Install only if you are comfortable with a retrospective tool that can save, load, and export JSON files at paths selected by the caller. Avoid pointing it at sensitive files or shared locations unless you intend to read or overwrite those files.

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
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding

The skill documents file persistence features such as save/load/export and includes example APIs that read from and write to disk, but it does not declare any tool scope like 'permissions' or 'allowed-tools'. That creates an authorization gap where a host agent may permit filesystem operations without explicit least-privilege boundaries, increasing the risk of unintended reads, overwrites, or data exfiltration if the skill is invoked on attacker-influenced paths.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The method writes retrospective contents to disk and creates directories as needed, but the only indication is a developer-facing docstring. There is no confirmation prompt, logging, or explicit user-facing disclosure at the point of execution, so users may not realize project retrospective data is being persisted to the filesystem.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

This method reads JSON from a filesystem path and registers the loaded retrospective in memory, but only a docstring describes the behavior. There is no visible disclosure, prompt, or logging to inform the user that local file contents are being accessed and incorporated into the analyzer state.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The method exports report contents to a JSON file and may create parent directories, but the operation is only described in a developer docstring. Without a visible warning, users may be unaware that analysis artifacts are being persisted to the local filesystem.

Content

No source excerpt is available for this finding.