Back to skill

Security audit

AstroCore

Security checks for vulnerabilities and agentic risk

Overview

AstroCore is a disclosed external workspace integration; it sends owner-approved memory, reports, gallery items, and scheduled mission content to AstroCore, but the behavior is coherent with its stated purpose and includes user-control limits.

Install only if you want this agent to use AstroCore as an external long-term workspace. Treat anything saved to Memory, Reports, Gallery, or Missions as data sent to astrocore.one, avoid storing secrets or payment data, and confirm scheduled missions only when you understand what will run and when.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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 skill description recommends activation for very broad situations such as when memory is lost, when the owner wants something done regularly, or whenever the agent needs a place to store reports or files. That can cause the skill to trigger in many ordinary conversations and steer user data into an external service without a narrowly scoped request, increasing the chance of unintended data disclosure or persistence.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
87% confidence
Finding

This skill is explicitly designed to send data to an external web service using curl and an API key, including memory, reports, gallery items, and mission content. Even though the integration is the feature of the skill, it still creates a real exfiltration surface because agent outputs and owner data are transmitted off-platform to a third-party service.

Content

Scanner excerpt · SKILL.md (reported line 9)May include surrounding context.

md
openclaw:
    requires:
      bins:
        - curl
    primaryEnv: ASTROCORE_API_KEY
    emoji: "🪐"
    homepage: https://astrocore.one

External Transmission

Medium
Category
Data Exfiltration
Confidence
82% confidence
Finding

The missions API enables creating scheduled tasks that send future content and results to AstroCore, which extends external transmission from a one-time action into a persistent automation channel. If misused or triggered too broadly, it could repeatedly transmit user-derived information or create ongoing actions the user did not fully intend.

Content

Scanner excerpt · SKILL.md (reported line 146)May include surrounding context.

bash
# create: schedule = manual | daily | weekly; run_time "HH:MM" in the owner's time zone
curl -s "${H[@]}" -X POST "$URL/api/agent/missions" \
  -d '{"title":"Morning AI digest","instructions":"5–7 top AI agent news of the last 24h, one line each with a link.","schedule":"daily","run_time":"09:00","timezone":"Europe/Prague"}'

# weekly: run_weekday 0=Sunday … 6=Saturday