Back to skill

Security audit

Incident Response Lifecycle

Security checks for vulnerabilities and agentic risk

Overview

This is a read-only incident-response coordination skill with one unnecessary SSH prerequisite that should be cleaned up but does not execute commands or access credentials.

This skill appears appropriate for teams that want structured incident coordination and post-mortem templates. Before installing, consider removing the unused SSH binary requirement so the declared runtime needs match the read-only documentation purpose.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Note
Location
SKILL.md:15
Finding
Unnecessary SSH Runtime Requirement Violates Least Privilege## Vulnerability Details **File Location**: `SKILL.md`, line 15 **Vulnerability Type**: Unnecessary privileged tooling declaration **Risk Level**: Low **Complete Code Snippet**: ```yaml metadata: safety: read-only author: network-security-skills-suite version: "1.0.0" openclaw: '{"emoji":"🚨","safetyTier":"read-only","requires":{"bins":["ssh"],"env":[]},"tags":["incident","nist","postmortem"],"mcpDependencies":[],"egressEndpoints":[]}' ``` ### Technical Analysis The Skill declares the `ssh` binary as a runtime prerequisite even though its documented function is limited to read-only incident-process coordination. The instructions explicitly defer network evidence collection and command execution to a separate Skill, and no reviewed file contains an SSH command or other operation requiring this binary. This mismatch violates least-privilege and capability-minimization principles. Declaring an unnecessary remote-access utility expands the tools available to the Skill runtime without a corresponding functional requirement. The declaration alone does not execute SSH, grant credentials, or create a remote session; therefore, this is a configuration-hardening issue rather than evidence of active compromise. ### Attack Path 1. The Skill is installed or loaded, and the runtime processes its declared prerequisites. 2. The runtime makes or verifies the availability of the `ssh` utility for this documentation-only Skill. 3. If the Skill instructions are later modified, combined with untrusted incident content, or otherwise influenced to invoke available tools, SSH could become an unnecessary remote-access execution channel. 4. Successful remote access would still require reachable infrastructure, suitable SSH options, and valid credentials or another independent authentication weakness. No direct exploitation path exists in the reviewed version because it contains no SSH invocation, credential access, destination host, ...[truncated 638 chars]
Remediation
## Remediation Suggestions 1. Remove the unnecessary SSH prerequisite by changing the requirement to an empty binary list: ```yaml openclaw: '{"emoji":"🚨","safetyTier":"read-only","requires":{"bins":[],"env":[]},"tags":["incident","nist","postmortem"],"mcpDependencies":[],"egressEndpoints":[]}' ``` 2. Keep the Skill documentation-only and continue delegating technical evidence collection to the separately scoped network incident-response Skill. 3. Add validation that rejects undeclared or unused binary requirements during packaging and review. 4. If SSH functionality is introduced later, document its exact purpose, constrain permitted destinations and commands, avoid implicit credential discovery, require host-key verification, and update the safety classification accordingly.
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Static analysis

No suspicious patterns detected.