Back to skill

Security audit

Research Paper Push

Security checks for vulnerabilities and agentic risk

Overview

This skill openly creates local research-paper alert subscriptions, queries OpenAlex, and registers OpenClaw cron jobs for scheduled summaries.

Before installing, confirm you want a local scheduled workflow: it will write subscription state under data/, query OpenAlex with your research topics, and create or remove named OpenClaw cron jobs when you add, update, or remove subscriptions. Do not put tokens or private channel credentials in the skill folder.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (10)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
91% confidence
Finding

The skill documentation instructs use of shell, network access, environment variables, and local file writes, but it declares no explicit tool scope or permission boundaries. That increases the chance an agent can invoke broader capabilities than intended, especially for cron registration and local state changes, without user-visible constraints.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The description states the skill will 'push concise daily Chinese summaries,' which imposes a specific language by default. Under the policy, forcing a language/locale without user opt-in is a natural-language policy violation unless a justified regional constraint is documented.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This line says the skill will 'format concise Chinese summaries for delivery,' presenting Chinese as a fixed output language. There is no accompanying opt-in, alternative language option, or justification that the skill is region-specific.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 120)May include surrounding context.

md
1. Builds English OpenAlex search queries from subscription topics.
2. Uses `from_publication_date` based on `--since`, saved `last_checked`, or `--days`.
3. Queries `https://api.openalex.org/works` sorted by recent publication date.
4. Converts OpenAlex works into paper records with title, authors, venue, date, DOI/link, citation count, and abstract.
5. Scores relevance against subscription topics plus common neuroimaging synonym groups.
6. Prefers requested journals/conferences but keeps highly relevant adjacent results.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The output style explicitly says default summaries are 'concise Chinese bullet points,' which sets a language constraint in the skill instructions. Because the file does not offer a user language choice or a documented compliance reason, this violates the language/locale policy.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/manage_papers.py (reported line 146)May include surrounding context.

python
# Some embedded Windows Python builds lack SSL support. Fall back to curl
    # when available so the skill can still query OpenAlex over HTTPS.
    try:
        result = subprocess.run(
            ["curl", "-L", "-sS", "--max-time", "30", "-A", "openclaw-research-paper-push/1.0", url],
            capture_output=True,
            text=True,

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill's stated purpose is managing research-paper subscriptions and summaries, but the implementation invokes external binaries (curl and openclaw) through subprocess.run. While querying OpenAlex and scheduling alerts are in scope, spawning OS-level commands is a broader capability than the manifest justifies and increases operational power beyond a typical paper-alert workflow.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This code file contains multiple hard-coded Chinese user-facing messages, beginning at L158 and continuing throughout the CLI flow, while the script description and arguments are otherwise in English. That creates a natural-language locale policy issue because users are not given any opt-in or configuration to choose the output language.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/manage_papers.py (reported line 349)May include surrounding context.

python
"--timezone", subscription.get("timezone") or "Asia/Shanghai",
        "--script", f"python scripts/manage_papers.py run --id {subscription['id']}",
    ]
    subprocess.run(cmd)
    print(f"⏰ Cron 任务已创建:{subscription['time']} ({subscription.get('timezone') or 'Asia/Shanghai'})")

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/manage_papers.py (reported line 355)May include surrounding context.

python
"--timezone", subscription.get("timezone") or "Asia/Shanghai",
        "--script", f"python scripts/manage_papers.py run --id {subscription['id']}",
    ]
    subprocess.run(cmd)
    print(f"⏰ Cron 任务已创建:{subscription['time']} ({subscription.get('timezone') or 'Asia/Shanghai'})")

Static analysis

No suspicious patterns detected.