Security audit

Nightly Workspace Audit

Security checks for vulnerabilities and agentic risk

Overview

This workspace-audit skill is not obviously malicious, but it can automatically scan and change or remove persistent workspace memory without clear approval or rollback controls.

Install only if you are comfortable with a skill that may read broad workspace and memory context and make persistent cleanup changes. Prefer running it manually in report-only mode, reviewing proposed deletions, moves, renames, and MEMORY.md edits before applying them, and keeping backups or version control for the workspace.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
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

Findings (3)

T01 · Skill Instruction Hijacking

Warning

Location: SKILL.md:57
Finding: Untrusted Local Override File Can Hijack Audit Instructions
Content: View full analysis
Remediation: View remediation

T09 · Insecure Skill Coding Practices

Warning

Location: SKILL.md:38
Finding: Automatic Destructive Workspace Cleanup Lacks Approval and Recovery Controls
Content: View full analysis
Remediation: View remediation

T05 · Unauthorized Access and Privilege Escalation

Note

Location: SKILL.md:19
Finding: Audit Procedure Performs Overly Broad Sensitive-Context Reconnaissance
Content: View full analysis
Remediation: View remediation

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

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The skill explicitly instructs the agent to merge, deduplicate, prune, archive, and rename files, including deleting entries deemed 'no longer useful,' but provides no explicit safety warning, approval gate, backup requirement, or dry-run mode. In an automated nightly or cron-triggered context, these destructive actions could silently cause data loss, corrupt user memory organization, or remove important information based on heuristic judgments.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The trigger language is extremely broad ('when the user asks to audit, clean up, or check workspace health') for a skill that performs wide filesystem scans and may modify, archive, rename, or delete files. This increases the chance of unintended invocation in ambiguous conversations, causing the agent to run destructive maintenance actions without sufficiently explicit user intent.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The file explicitly instructs the agent to remove sections from MEMORY.md, move files to memory/archive/, and update references, but it does not require an explicit user-facing warning or confirmation about the risk of altering or effectively deleting stored memory content. In a workspace-audit skill that may run automatically on a nightly cron job, these retention changes can silently hide, relocate, or remove important context and create data-loss or integrity issues if the classification is wrong.
Content: No source excerpt is available for this finding.