Back to skill

Security audit

Human-Agent Collaboration Protocol

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Feishu meeting-prep and follow-up workflow with confirmation before external document or task changes.

Install only if you want an agent to help prepare and follow up on meetings in Feishu. Review generated Context Cards and Outputs carefully before confirming, because confirmation can create Feishu docs, update docs, create assigned tasks, and save meeting output to local memory.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The trigger examples are broad enough to activate on ordinary conversation about meeting or talking to someone, which can cause the skill to run when the user did not explicitly request this workflow. In this skill’s context, over-triggering can lead to unsolicited questioning and preparation of shared documents or downstream enterprise actions tied to meetings, increasing the chance of inappropriate automation around sensitive interpersonal or business discussions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The manifest description is written as a mandatory Chinese-only instruction set and does not indicate any user choice of language or locale. Under the policy, language constraints should be optional or explicitly justified when region-specific.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The catch-all scope of 'any scenario where I want to talk to XXX about XXX' lacks clear boundaries and can sweep in casual, personal, or sensitive conversations that do not warrant automated workflow support. Because this skill later instructs the agent to create shared Context Cards and process post-meeting notes, ambiguous activation raises the risk of privacy violations, unnecessary data handling, and mistaken enterprise-tool actions.

Content

No source excerpt is available for this finding.