Security audit

DashClaw Platform Intelligence

Security checks for vulnerabilities and agentic risk

Overview

This DashClaw skill is a coherent platform reference and diagnostic helper, with expected API-key use and opt-in write checks.

Install this only if you use DashClaw and intend to let the skill query your DashClaw instance. Prefer a least-privilege or test API key, review commands before using --full, --validate, live bootstrap, /api/doctor/fix, or --capture-setup-proof, and avoid running these against production unless the resulting test records and diagnostic uploads are acceptable.

SkillSpector

Vulnerability Patterns

MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
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 (1)

Description-Behavior Mismatch

Medium

Confidence: 89% confidence
Finding: The script is presented as an integration validator, but it can perform state-changing POST/PATCH requests and optionally submit 'setup proof' to a remote endpoint. That behavior exceeds passive verification and can create or modify server-side data unexpectedly, which is risky in troubleshooting or audit contexts where users may assume the tool is read-only unless they inspect the code carefully.

Static analysis

Detected: suspicious.dangerous_exec, suspicious.env_credential_access

Shell command execution detected (child_process).

Critical

Code: suspicious.dangerous_exec
Location: scripts/bootstrap-agent-quick.mjs:108

Environment variable access combined with network send.

Critical

Code: suspicious.env_credential_access
Location: scripts/bootstrap-agent-quick.mjs:48

Environment variable access combined with network send.

Critical

Code: suspicious.env_credential_access
Location: scripts/diagnose.mjs:37

Environment variable access combined with network send.

Critical

Code: suspicious.env_credential_access
Location: scripts/validate-integration.mjs:36