Security audit

Agent Knowledge Capture

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward knowledge-capture helper, but users should understand that it stores content persistently and recommends automated cleanup that can modify stored files.

Install only if you want an agent-readable local knowledge store. Treat `know add` as permanent capture of URLs, summaries, tags, and research notes in the configured directory, and review that directory before storing sensitive material. Do not enable heartbeat or cron `know tidy --fix` unless the knowledge directory is dedicated to this tool and you are comfortable with automatic file moves, normalization, and removal of empty entries.

A.I.G

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

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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 (3)

Missing User Warnings

High

Confidence: 97% confidence
Finding: The skill recommends automated execution of 'know tidy --fix' in heartbeats or cron while describing behavior such as moving misplaced files, normalizing content, and removing empty items. Automated background modification of a persistent knowledge store without an explicit safety warning creates a real risk of unintended file changes, data loss, or corruption, especially if the configured directory is wrong or contains mixed-use content.

Vague Triggers

Medium

Confidence: 93% confidence
Finding: The manifest description is extremely broad ('use when saving anything worth re-finding later' and related retrieval phrasing), which can cause the agent to invoke this skill for many ordinary research, note-taking, or recall tasks. Because the skill performs persistent file writes, broad routing language increases the chance of unintended data capture and filesystem modification beyond what the user explicitly requested.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The add commands are presented as routine operations, but the description does not clearly warn that they create markdown files with metadata and update a persistent INDEX.md under a user directory. This can lead to silent persistence of URLs, summaries, research outputs, or other potentially sensitive content when a user may expect an ephemeral action.

Static analysis

No suspicious patterns detected.