Back to skill

Security audit

my-skill-name

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed helper for creating, editing, validating, and packaging skills, with no hidden network, credential, or background persistence behavior found.

Install this only if you want an agent to help create or modify skills. Review diffs to SKILL.md and bundled resources before packaging or installing generated skills, and avoid pointing the packaging script at directories containing unrelated private files.

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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Rogue AgentSelf-Modification, Session Persistence
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (13)

Tp4

High
Category
MCP Tool Poisoning
Confidence
93% confidence
Finding

The code only supports one narrow subset of the declared description: creating a new skill scaffold. It normalizes a skill name, creates a directory, writes a template SKILL.md, optionally creates resource folders, and adds example files. There is no logic for opening or modifying an existing skill, auditing content, reviewing compliance, removing stale files, moving files between directories, or validating against any spec. Because the declared purpose prominently includes edit/improve/audit/cleanup/restructure/validate capabilities that are absent from the code, the description materially overstates the behavior of this code chunk.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The declared description emphasizes authoring, editing, reviewing, cleaning up, restructuring, and auditing skills. The supplied code does not implement those functions. Instead, its primary behavior is to validate a skill folder and package it into a .skill archive file on disk. While validation is loosely adjacent to auditing/spec compliance, the dominant capability here is packaging/distribution, which is not declared. This is a materially different primary purpose, so it should be flagged as a mismatch.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The declared description says this skill is for authoring, editing, improving, auditing, or restructuring AgentSkills and SKILL.md content. However, the actual code chunk is not implementing skill creation or modification logic; it is a unit/regression test module focused on the behavior of a package_skill function. Its primary purpose is to test packaging security properties—filesystem traversal, symlink exclusion, archive contents, and output-archive handling. That is a materially different primary purpose from skill authoring or cleanup, so this should be flagged as a mismatch.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The declared description presents a user-facing skill for authoring, editing, reviewing, tidying, and auditing AgentSkills directories and SKILL.md files. The actual code shown does not implement those workflows. Instead, it is a test suite for a validation component, specifically checking acceptance/rejection of SKILL.md frontmatter variants and fallback parsing without PyYAML. While validation is mentioned in the description, this code is not performing skill editing/restructuring/auditing itself; it is supporting developer-side automated testing of another module. That is a materially different primary purpose from the declared skill behavior.

Content

No source excerpt is available for this finding.

Self-Modification

High
Category
Rogue Agent
Confidence
92% confidence
Finding

This skill explicitly instructs the agent to create and modify skill files, including writing SKILL.md and bundled resources. In an agent ecosystem, self-modification is dangerous because a compromised prompt, malicious repository content, or confused trigger can cause the agent to alter its own capabilities, persistence, or future behavior without an independent approval boundary.

Content

Scanner excerpt · SKILL.md (reported line 208)May include surrounding context.

md
1. Understand the skill with concrete examples
2. Plan reusable skill contents (scripts, references, assets)
3. Initialize the skill (run init_skill.py)
4. Edit the skill (implement resources and write SKILL.md)
5. Package the skill (run package_skill.py)
6. Iterate based on real usage

Self-Modification

High
Category
Rogue Agent
Confidence
92% confidence
Finding

The 'Update SKILL.md' instruction is another explicit self-modification pathway. Since SKILL.md controls how and when a skill triggers, permitting the agent to rewrite it can silently expand scope, alter future behavior, or embed unsafe instructions that persist across sessions, making this more dangerous than ordinary content editing.

Content

Scanner excerpt · SKILL.md (reported line 315)May include surrounding context.

md
If you used `--examples`, delete any placeholder files that are not needed for the skill. Only create resource directories that are actually required.

#### Update SKILL.md

**Writing Guidelines:** Always use imperative/infinitive form.

Self-Modification

High
Category
Rogue Agent
Confidence
85% confidence
Finding

Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Content

Scanner excerpt · scripts/init_skill.py (reported line 307)May include surrounding context.

python
# Print next steps
    print(f"\n[OK] Skill '{skill_name}' initialized successfully at {skill_dir}")
    print("\nNext steps:")
    print("1. Edit SKILL.md to complete the TODO items and update the description")
    if resources:
        if include_examples:
            print("2. Customize or delete the example files in scripts/, references/, and assets/")

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/test_package_skill.py (reported line 67)May include surrounding context.

python
def test_skips_symlink_to_external_file(self):
        skill_dir = self.create_skill("symlink-file-skill")
        outside = self.temp_dir / "outside-secret.txt"
        outside.write_text("super-secret\n")
        link = skill_dir / "loot.txt"
        out_dir = self.temp_dir / "out"

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/test_package_skill.py (reported line 92)May include surrounding context.

python
def test_skips_symlink_to_external_file(self):
        skill_dir = self.create_skill("symlink-file-skill")
        outside = self.temp_dir / "outside-secret.txt"
        outside.write_text("super-secret\n")
        link = skill_dir / "loot.txt"
        out_dir = self.temp_dir / "out"

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/test_package_skill.py (reported line 109)May include surrounding context.

python
def test_skips_symlink_to_external_file(self):
        skill_dir = self.create_skill("symlink-file-skill")
        outside = self.temp_dir / "outside-secret.txt"
        outside.write_text("super-secret\n")
        link = skill_dir / "loot.txt"
        out_dir = self.temp_dir / "out"

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill description and trigger phrases are broad enough to overlap with many generic requests such as reviewing, cleaning up, or improving content. This can cause the skill to activate in contexts outside skill authoring, steering the agent toward inappropriate workflows, unnecessary file operations, or unintended modifications; in a meta-skill that can create/edit other skills, misfires are more dangerous because it has authority over packaging, restructuring, and validation actions.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill uses very broad trigger phrases such as 'review', 'audit', 'clean up', and 'improve', which can cause it to activate for many unrelated requests. Because this skill instructs the agent to create, edit, delete placeholders, run scripts, and package artifacts, overbroad triggering can route ordinary review tasks into a self-modifying or filesystem-writing workflow, increasing the chance of unintended changes.

Content

No source excerpt is available for this finding.