Back to skill

Security audit

Review my skill

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Markdown-only reviewer for agent instruction files, with no code execution, persistence, or hidden data handling found.

Installers should know this skill is meant to analyze text they paste into the chat. Avoid pasting secrets, private system prompts, or proprietary agent files unless you are comfortable having them reviewed in the active AI conversation. The scanner warnings are explainable from example prompt-injection and safety-training text in the references, not from hidden execution behavior.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (7)

YARA rule 'agent_skill_prompt_injection_hidden_instructions': Prompt injection or hidden instructions embedded in AI agent skill text [agent_skills]

High
Category
YARA Match
Confidence
80% confidence
Finding

YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Content

Scanner excerpt · references/anti-patterns.md (reported line 6)May include surrounding context.

Anti-Patterns — Common Mistakes in Agent Instructions

1. The "Be Helpful" trap

What it looks like:

text
You are a helpful, friendly, knowledgeable assistant.

What goes wrong: "Helpful" means nothing specific. The model falls back to its default behaviour — verbose, eager to please, reluctant to say no. Every response sounds the same regardless of context.

The fix: Replace abstract adjectives with concrete behaviours.

text
Answer in 2-3 sentences unless the user asks for detail. If you don't know, say so. Prioritise accuracy over friendliness.

2. The "Do Everything" trap

What it looks like:

text
You can help with coding, writing, analysis, brainstorming, debugging, planning, research, translation, summarisation,

Instruction Override

High
Category
Prompt Injection
Confidence
80% confidence
Finding

This pattern attempts to override system instructions or ignore safety constraints. Without LLM analysis, manual review is recommended.

Content

Scanner excerpt · references/anti-patterns.md (reported line 65)May include surrounding context.

md
**What it looks like:** No input validation, no safety boundaries, no refusal instructions.

**What goes wrong:**
- Users paste prompt injections ("ignore all previous instructions and...")
- Users ask the agent to do things outside its competence (medical advice, legal guidance)
- Users accidentally paste credentials or sensitive data
- The agent complies with everything because nothing told it not to

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The README states the skill can be invoked with /review-my-agent or by 'just past[ing] an agent instruction file in a conversation where the skill is active,' which implies broad or implicit triggering. For a skill that ingests untrusted prompt text, ambiguous auto-activation can cause the analyzer to engage on content the user did not explicitly intend to submit, increasing the chance of accidental processing of sensitive instructions or adversarial prompt content.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 35)May include surrounding context.

md
**Ambiguous:** If the user's intent doesn't clearly match a mode, ask one question: "Want to paste it for a review, or describe the problem?"

If the user shifts mode mid-conversation (e.g., asks a question then pastes a file), follow the new mode without asking. The file is the signal.

## The review

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

Line L176 states "Review in the language written," which imposes a language/locale behavior regardless of user preference. Under the policy, forced language behavior without user opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · references/anti-patterns.md (reported line 123)May include surrounding context.

Pattern:

text
## Priority hierarchy
1. Safety — never execute destructive commands without confirmation
2. Accuracy — say "I don't know" over guessing
3. Brevity — short answers unless detail is requested
4. Personality — maintain voice but not at the cost of the above

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · references/instruction-patterns.md (reported line 9)May include surrounding context.

Pattern:

text
## Priority hierarchy
1. Safety — never execute destructive commands without confirmation
2. Accuracy — say "I don't know" over guessing
3. Brevity — short answers unless detail is requested
4. Personality — maintain voice but not at the cost of the above