Security audit

insight

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent note-analysis tool, but it can broadly read sensitive personal notes and defaults to modifying notes without enough upfront warning.

Install only if you are comfortable letting an agent process private notes. Prefer `--output reply-only`, avoid `--source all` unless needed, scope directories narrowly, and do not run it on journals, health notes, relationship details, or third-party information without understanding where your agent or model may send and store that content.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (3)

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The README explicitly promotes analysis of personal notes and mental-health-related content, including 'Self-Discovery' and 'Emotional & Mental Health' perspectives, but provides no warning about handling sensitive personal data, privacy expectations, or limits of the analysis. This can lead users to submit highly sensitive journals, therapy-like content, or third-party information to an LLM-based skill without informed consent about confidentiality, retention, or misinterpretation risks.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The skill advertises itself as a note analysis tool but does not clearly warn in its top-level description that some output modes will append to existing notes or create new files. In an agent setting, this can lead users to invoke the skill expecting read-only behavior, causing unintended modification of personal knowledge bases and possible corruption, clutter, or propagation of sensitive analysis into synced note stores.

Missing User Warnings

Medium

Confidence: 87% confidence
Finding: The skill supports searching across multiple personal note sources and mentions token/API-based access, but it provides no privacy warning about aggregating sensitive data from Obsidian, Flomo, Evernote, Dedao, or generic directories. This increases the risk that users unknowingly expose private notes, credentials, relationship data, or mental-health-related content to the agent or external integrations without informed consent.

Static analysis

No suspicious patterns detected.