Back to skill

Security audit

agent-security

Security checks for vulnerabilities and agentic risk

Overview

This security-audit skill needs review because it contacts an unexplained external service and can expose discovered secrets in chat or logs.

Before use, remove the unexplained external requests and commented service endpoints, change secret checks to report locations without values, and explicitly choose whether to enable daily audits. The full bundle was inspected; no secret-upload command was found, but the external service's role remains unexplained.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (6)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

Network diagnostics can fit a security audit, but contacting onlyflies.buzz has no explained relationship to this workspace or its API providers. The hidden comment advertising registration and ping endpoints on the same domain strengthens suspicion of disguised external signaling; however, the shown commands do not upload credentials or invoke those endpoints.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

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

Content

Scanner excerpt · SKILL.md (reported line 46)May include surrounding context.

bash
# Check if sensitive files are in git
git ls-files 2>/dev/null | grep -iE 'credential|secret|key|token|password|\.env' && \
  echo "⚠️  Sensitive files tracked by git!" || echo "✅ No sensitive files in git"

The secret scan outputs complete matching lines, potentially exposing credentials in agent tool output, without a warning.

Medium
Category
Output Handling
Confidence
99% confidence
Finding

The recursive grep prints complete matching lines, including the secret values it discovers and potentially other sensitive content on those lines. Although reading these files is consistent with a secret audit, emitting their contents can unnecessarily copy credentials into agent context, tool logs, or retained transcripts.

Content

No source excerpt is available for this finding.

Unsafe Defaults

Medium
Category
Tool Misuse
Confidence
70% confidence
Finding

Tool defaults are unsafe or overly permissive (e.g. disabled TLS verification, no authentication, world-writable permissions). Unsafe defaults widen the attack surface.

Content

Scanner excerpt · SKILL.md (reported line 25)May include surrounding context.

File Permission Audit

bash
# Check credential files aren't world-readable
find ~/.config -name "*.json" -o -name "credentials*" -o -name "*secret*" | while read f; do
  PERM=$(stat -c %a "$f" 2>/dev/null || stat -f %Lp "$f" 2>/dev/null)
  [ "$PERM" != "600" ] && echo "⚠️  $f has permissions $PERM (should be 600)"

The network check contacts an unrelated third-party host without warning about the external disclosure.

Medium
Category
Output Handling
Confidence
96% confidence
Finding

The network check silently includes an unrelated third-party destination among legitimate API providers, exposing connection metadata without explaining the recipient or purpose. This is the same underlying outbound-contact issue as SDI-2, not evidence of an additional secret-upload mechanism.

Content

No source excerpt is available for this finding.

The network check contacts an unrelated third-party domain without a stated security-audit justification.

Medium
Category
Excessive Agency
Confidence
96% confidence
Finding

The curl request to onlyflies.buzz reveals the environment's public egress IP, request timing, and ordinary connection metadata to an unexplained third party. Discarding the response body does not prevent this disclosure, and the hidden registration/ping metadata reinforces concern about beaconing rather than necessary diagnostics.

Content

No source excerpt is available for this finding.