ClawHub
Back to skill

Security audit

Awesome Paper Skill

Security checks for vulnerabilities and agentic risk

Overview

The skill does what it advertises, but it can publish to and change GitHub repositories with insufficient safeguards.

Review before installing. Use it only with an explicit repository owner/name, inspect the generated README before publishing, and avoid running the publish step unless you are comfortable with it creating or updating a GitHub repo and changing visibility through your authenticated gh session.

A.I.G

Vulnerability Patterns
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • 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
Findings (2)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/build_awesome_readme.py:91
Finding

Unsanitized External Publication Metadata Injected into Generated Markdown

Content
View full analysis
Remediation
View remediation

T05 · Unauthorized Access and Privilege Escalation

Note
Location
scripts/publish_repo.py:17
Finding

Existing Repository Visibility Is Changed Without an Explicit Visibility Request

Content
View full analysis
Remediation
View remediation

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (10)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill directs the agent to create or update a repository and push changes, but it does not require a clear user-facing warning or confirmation immediately before performing those external write actions. This is dangerous because it can cause unintended modification of a user's GitHub repository, including creating new repos or overwriting README content, especially when combined with the broad invocation logic and possible default repo selection.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill is invoked for a broad 'full pipeline' request and allows missing owner/repo values to fall back to workspace defaults, which can cause the agent to perform repository-affecting actions without a narrowly scoped, explicit user confirmation. In this context, the broad trigger is more dangerous because the pipeline culminates in creating or updating and pushing to GitHub, so accidental invocation can lead to unintended external side effects.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The formatting rules require 'English-only,' which imposes a language constraint regardless of the user's preferences. The file does not offer a language choice or explain a justified locale-specific reason for this restriction.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% 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 · scripts/fetch_papers.py (reported line 55)May include surrounding context.

python
def fetch_crossref(topic: str, limit: int):
    q = urllib.parse.quote(topic)
    url = f"https://api.crossref.org/works?query={q}&rows={limit}&sort=published&order=desc"
    txt = http_get(url)
    data = json.loads(txt)
    items = data.get("message", {}).get("items", [])

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% 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 · scripts/fetch_papers.py (reported line 101)May include surrounding context.

python
def fetch_semantic(topic: str, limit: int):
    q = urllib.parse.quote(topic)
    fields = "title,abstract,year,venue,url,externalIds,publicationTypes"
    url = f"https://api.semanticscholar.org/graph/v1/paper/search?query={q}&limit={limit}&fields={fields}"
    txt = http_get(url)
    data = json.loads(txt)
    out = []

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/publish_repo.py (reported line 11)May include surrounding context.

python
def run(cmd, cwd=None):
    subprocess.run(cmd, cwd=cwd, check=True)


def main():

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/publish_repo.py (reported line 34)May include surrounding context.

python
run(["git", "add", "README.md"], cwd=td)
        run(["git", "commit", "-m", "docs: initialize awesome papers README"], cwd=td)

        exists = subprocess.run(["gh", "repo", "view", repo], stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL).returncode == 0
        if not exists:
            run(["gh", "repo", "create", repo, vis_flag, "--source", td, "--remote", "origin", "--push"], cwd=td)
        else:

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The script can create a GitHub repository, configure a remote, push content, and change repository visibility without any explicit user confirmation or dry-run safeguard. In an agent or automation context, that makes unintended publication or accidental modification of remote resources much more likely, especially because the action targets authenticated GitHub state and may expose data publicly if --visibility public is chosen.

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/publish_repo.py (reported line 38)May include surrounding context.

python
if not exists:
            run(["gh", "repo", "create", repo, vis_flag, "--source", td, "--remote", "origin", "--push"], cwd=td)
        else:
            remotes = subprocess.check_output(["git", "remote"], cwd=td).decode()
            if "origin" not in remotes:
                run(["git", "remote", "add", "origin", f"https://github.com/{repo}.git"], cwd=td)
            run(["git", "push", "-u", "origin", "main"], cwd=td)

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

This code file performs a file write to a user-specified path, which is a safety-relevant operation under the rule. While it prints after completion, there is no prior disclosure, confirmation prompt, or inline comment/docstring warning that the target file will be overwritten if it already exists.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.