Security audit

Dont Waste Food

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent cooking helper that stores recipe sessions locally, with some privacy and deletion caveats users should understand.

Install only if you are comfortable with the skill saving cooking history and food images in its local workspace. Use /clear carefully because it deletes the skill's saved history, stats, and workspace data without an extra confirmation step.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands

Findings (5)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 94% confidence
Finding: The skill instructs the agent to read and write local files and to fetch remote images, but it declares no permissions or user-facing consent boundaries. That mismatch is dangerous because users and the hosting platform may not realize the skill can persist session data locally and make network requests, increasing the risk of unauthorized data handling or SSRF-style abuse through attacker-supplied image URLs.

Context-Inappropriate Capability

Medium

Confidence: 89% confidence
Finding: The module exposes a command-line `clear` action that calls `delete_all()`, which recursively removes the entire skill workspace. While this appears intended as a maintenance/reset feature, it is broader than ordinary session handling and creates a destructive capability that can erase all persisted user data if invoked accidentally or by another component.

Vague Triggers

Medium

Confidence: 86% confidence
Finding: The trigger phrases are broad enough that ordinary mentions of leftovers, fridge ingredients, or unused food could invoke the skill when the user did not intend to activate it. In this skill, unintended invocation matters because activation can lead to image handling, safety questioning, and local/session persistence flows, causing unexpected processing of user content.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The skill clearly describes downloading images and storing sessions/history in workspace files, but it does not clearly disclose retention, deletion behavior, or obtain informed consent before saving user inputs. This is a privacy and data-governance issue because food photos, notes, timestamps, and cooking history can reveal household habits and may be retained longer than users expect.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: `delete_all()` performs irreversible deletion with `shutil.rmtree(WORKSPACE)` and immediately recreates the directory, but there is no confirmation prompt, dry-run mode, backup, or authorization check. This makes accidental invocation or misuse by connected tooling likely to cause total loss of the user's cooking history and stats.

Static analysis

No suspicious patterns detected.