ClawHub
Back to skill

Security audit

JWT

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only JWT security guidance skill with no hidden execution, credential access, persistence, or install-time behavior.

This skill is reasonable to install as a JWT security checklist. Users should still verify library-specific settings against their chosen JWT library and only run attack-test examples against systems they own or are authorized to test.

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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

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 · SKILL.md (reported line 48)May include surrounding context.

md
## Token Lifecycle

- Access tokens: short-lived (5-15 min)—limits damage if stolen
- Refresh tokens: longer-lived, stored securely—used only to get new access tokens
- Refresh token rotation: issue new refresh token on each use, invalidate old one
- Revocation is hard—JWTs are stateless; use short expiry + refresh, or maintain blacklist

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 · SKILL.md (reported line 49)May include surrounding context.

md
## Token Lifecycle

- Access tokens: short-lived (5-15 min)—limits damage if stolen
- Refresh tokens: longer-lived, stored securely—used only to get new access tokens
- Refresh token rotation: issue new refresh token on each use, invalidate old one
- Revocation is hard—JWTs are stateless; use short expiry + refresh, or maintain blacklist

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 · lifecycle.md (reported line 12)May include surrounding context.

md
## Refresh Token Traps

- Long-lived refresh token = high value target—store more securely than access token
- Refresh token rotation: issue new on each use, invalidate old
- Stolen refresh token without rotation = unlimited access
- Refresh endpoint should be rate-limited—attacker tries stolen token

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 · validation.md (reported line 5)May include surrounding context.

md
## Signature Verification Traps

- Decode without verify = reading claims without checking authenticity—useless for auth
- Verify with wrong key = invalid signature for valid token—check key first
- Algorithm must be enforced server-side—never trust token's `alg` claim
- Empty signature with `alg: none`—reject explicitly

Static analysis

No suspicious patterns detected.