ClawHub
Back to skill

Security audit

Cnbc Geopolitics Fetcher

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly does what it says, but it ships a real-looking Discord webhook credential and tells agents to use it for external posting.

Review before installing. Do not use the bundled webhook; assume it is compromised, revoke or replace it, and provide your own webhook through a safer local secret mechanism. Run in an isolated Python environment and expect the skill to send scraped CNBC article content to Discord when invoked.

A.I.G

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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

Error
Location
references/config.md:1
Finding

Hardcoded Discord Webhook Credential

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
SKILL.md:143
Finding

Unpinned Third-Party Dependency Installation

Content
View full analysis
Remediation
View remediation
requests== scrapling== ``` 2. Generate and verify cryptographic hashes for all direct and transitive dependencies. 3. Install with hash enforcement: ```bash python -m pip install --require-hashes -r requirements.txt ``` 4. Use a lock-file tool that records the complete transitive dependency graph. 5. Perform installation inside a dedicated, non-privileged virtual environment or container. 6. Avoid installing packages as an administrator or root user. 7. Add automated dependency vulnerability and provenance scanning to the release process. 8. Review dependency updates before changing pinned versions rather than automatically consuming the newest release. ]]>

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (34)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The documentation includes full Discord webhook URLs that appear live, which are effectively bearer secrets. Anyone who obtains them can post arbitrary messages to the associated Discord channel, abuse the integration, or use them as a covert exfiltration destination.

Content

No source excerpt is available for this finding.

Ssd 3

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

Embedding a live-looking webhook directly in examples encourages credential reuse and makes the secret immediately available to any reader or automated system parsing the document. This materially increases the chance of unauthorized posting, spam, channel compromise, and use of the webhook as an exfiltration sink.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The file exposes a live Discord webhook URL directly in repository content, which is effectively a secret that allows unauthenticated posting to a Discord channel. Anyone who obtains this URL can spam the channel, send misleading messages, abuse integrations, or force the webhook owner to rotate and remediate the credential.

Content

No source excerpt is available for this finding.

Unbounded Output

Medium
Category
Output Handling
Confidence
60% confidence
Finding

Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.

Content

Scanner excerpt · SKILL.md (reported line 4)May include surrounding context.

md
# CNBC Geopolitics Fetcher Skill

## Purpose
Fetches latest CNBC geopolitical news articles, extracts complete factual data (NO truncation), and posts each article individually to Discord with enhanced analyst/Polymarket-relevant context.

## What This Skill Does
1. **Scrapes** 5 latest CNBC articles on geopolitics (Iran, Middle East, oil markets, energy policy)

Unbounded Output

Medium
Category
Output Handling
Confidence
60% confidence
Finding

Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.

Content

Scanner excerpt · SKILL.md (reported line 8)May include surrounding context.

md
# CNBC Geopolitics Fetcher Skill

## Purpose
Fetches latest CNBC geopolitical news articles, extracts complete factual data (NO truncation), and posts each article individually to Discord with enhanced analyst/Polymarket-relevant context.

## What This Skill Does
1. **Scrapes** 5 latest CNBC articles on geopolitics (Iran, Middle East, oil markets, energy policy)

Unbounded Output

Medium
Category
Output Handling
Confidence
60% confidence
Finding

Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.

Content

Scanner excerpt · SKILL.md (reported line 81)May include surrounding context.

md
# CNBC Geopolitics Fetcher Skill

## Purpose
Fetches latest CNBC geopolitical news articles, extracts complete factual data (NO truncation), and posts each article individually to Discord with enhanced analyst/Polymarket-relevant context.

## What This Skill Does
1. **Scrapes** 5 latest CNBC articles on geopolitics (Iran, Middle East, oil markets, energy policy)

Unbounded Output

Medium
Category
Output Handling
Confidence
60% confidence
Finding

Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.

Content

Scanner excerpt · SKILL.md (reported line 196)May include surrounding context.

md
# CNBC Geopolitics Fetcher Skill

## Purpose
Fetches latest CNBC geopolitical news articles, extracts complete factual data (NO truncation), and posts each article individually to Discord with enhanced analyst/Polymarket-relevant context.

## What This Skill Does
1. **Scrapes** 5 latest CNBC articles on geopolitics (Iran, Middle East, oil markets, energy policy)

Unbounded Output

Medium
Category
Output Handling
Confidence
60% confidence
Finding

Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.

Content

Scanner excerpt · SKILL.md (reported line 216)May include surrounding context.

md
# CNBC Geopolitics Fetcher Skill

## Purpose
Fetches latest CNBC geopolitical news articles, extracts complete factual data (NO truncation), and posts each article individually to Discord with enhanced analyst/Polymarket-relevant context.

## What This Skill Does
1. **Scrapes** 5 latest CNBC articles on geopolitics (Iran, Middle East, oil markets, energy policy)

Unbounded Output

Medium
Category
Output Handling
Confidence
60% confidence
Finding

Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.

Content

Scanner excerpt · scripts/fetch_cnbc_geopolitics.py (reported line 8)May include surrounding context.

python
# CNBC Geopolitics Fetcher Skill

## Purpose
Fetches latest CNBC geopolitical news articles, extracts complete factual data (NO truncation), and posts each article individually to Discord with enhanced analyst/Polymarket-relevant context.

## What This Skill Does
1. **Scrapes** 5 latest CNBC articles on geopolitics (Iran, Middle East, oil markets, energy policy)

Unbounded Output

Medium
Category
Output Handling
Confidence
60% confidence
Finding

Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.

Content

Scanner excerpt · scripts/format_new.py (reported line 14)May include surrounding context.

python
# CNBC Geopolitics Fetcher Skill

## Purpose
Fetches latest CNBC geopolitical news articles, extracts complete factual data (NO truncation), and posts each article individually to Discord with enhanced analyst/Polymarket-relevant context.

## What This Skill Does
1. **Scrapes** 5 latest CNBC articles on geopolitics (Iran, Middle East, oil markets, energy policy)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill is designed to fetch external content and send it to a Discord webhook, but the documentation does not clearly foreground that this transmits data to a third-party service. In an agent setting, that omission increases the risk of unintended exfiltration of fetched content or operator-provided context to an external endpoint.

Content

No source excerpt is available for this finding.

Unbounded Output

Medium
Category
Output Handling
Confidence
60% confidence
Finding

Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.

Content

Scanner excerpt · SKILL.md (reported line 100)May include surrounding context.

md
- Extracts: title, full article text, description metadata
- Falls back to regex extraction if Scrapling fails

### 3. Fact Extraction (NO Truncation)
- Splits article text into complete sentences
- Scans for 6 fact categories:
  1. **Official**: Quoted statements with attribution (said/stated/announced)

Unbounded Output

Medium
Category
Output Handling
Confidence
75% confidence
Finding

Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.

Content

Scanner excerpt · SKILL.md (reported line 113)May include surrounding context.

md
### 4. Market Impact Extraction
- Scans for energy, stocks, currency, analyst, Polymarket sentences
- Returns complete sentences (NO length limits)
- Categories: Energy, Stocks, Currency, Analyst, Polymarket

### 5. Discord Posting

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The customization section instructs users to run the script with an --output argument to write results to output.md, but elsewhere the skill is consistently described as scraping CNBC articles and posting them to Discord. This is an active documentation-level divergence because it presents a concrete alternate behavior not supported by the stated skill intent in the rest of the file.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill explicitly instructs an LLM/agent to read a local config file for a webhook secret and use it to transmit data externally. That behavior normalizes secret retrieval and outbound posting without an approval boundary, which is risky in agentic environments where local files may contain sensitive credentials.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

For plain text files, vague or missing trigger descriptions are in scope. This file provides article summaries only and does not define when the skill should activate, what phrases should invoke it, or any exclusions, leaving invocation behavior ambiguous.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

This code performs outbound HTTP POST requests to a Discord webhook, transmitting scraped article content and URLs off-system. While the module docstring mentions posting to Discord, the runtime path that actually sends data provides no confirmation prompt or explicit user-facing warning at the point of transmission, which is the kind of disclosure this rule looks for in code files.

Content

No source excerpt is available for this finding.

Unbounded Output

Medium
Category
Output Handling
Confidence
86% confidence
Finding

The script intentionally preserves full sentences from untrusted remote HTML and later incorporates them into Discord messages without effective size bounds. A malicious or malformed page could supply extremely long 'sentences,' causing oversized payloads, repeated failures, noisy output, or downstream denial-of-service behavior in the posting workflow.

Content

Scanner excerpt · scripts/fetch_cnbc_geopolitics.py (reported line 274)May include surrounding context.

python
# Split into sentences
    sentences = re.split(r'(?<=[.!?])\s+', text)
    
    # Find oil/energy sentences - NO length limit
    for sent in sentences:
        if any(x in sent.lower() for x in ['oil', 'crude', 'WTI', 'Brent', 'barrel', 'energy']):
            if len(sent) > 20:  # Only minimum length check

Unbounded Output

Medium
Category
Output Handling
Confidence
86% confidence
Finding

This branch appends full stock-related sentences from scraped content with no upper bound, contributing to unbounded message construction from untrusted input. In this skill context, remote publishers control the source text, so malformed or adversarially long text can degrade reliability and create output flooding or repeated posting errors.

Content

Scanner excerpt · scripts/fetch_cnbc_geopolitics.py (reported line 281)May include surrounding context.

python
impacts.append(f"Energy: {sent}")
                break
    
    # Find stock sentences - NO length limit
    for sent in sentences:
        if any(x in sent.lower() for x in ['S&P', 'NASDAQ', 'Dow', 'Stoxx', 'stock', 'market']):
            if len(sent) > 20:

Unbounded Output

Medium
Category
Output Handling
Confidence
86% confidence
Finding

Currency-related sentences are copied verbatim from external content without a maximum length, creating the same unbounded output condition. Because the data comes from the network and is later sent to another network sink, this can amplify oversized or malformed input into operational failures.

Content

Scanner excerpt · scripts/fetch_cnbc_geopolitics.py (reported line 288)May include surrounding context.

python
impacts.append(f"Stocks: {sent}")
                break
    
    # Find currency sentences - NO length limit
    for sent in sentences:
        if any(x in sent.lower() for x in ['dollar', 'EUR', 'USD', 'currency', 'exchange']):
            if len(sent) > 20:

Unbounded Output

Medium
Category
Output Handling
Confidence
86% confidence
Finding

Analyst/economic forecast snippets are accepted with only a minimum-length check and no maximum bound, so a single long sentence can balloon the outgoing message. In an automated posting pipeline, this can trigger repeated exceptions, partial delivery, or excessive resource use.

Content

Scanner excerpt · scripts/fetch_cnbc_geopolitics.py (reported line 295)May include surrounding context.

python
impacts.append(f"Currency: {sent}")
                break
    
    # Find analyst/economic forecast sentences - NO length limit
    for sent in sentences:
        if any(x in sent.lower() for x in ['analyst', 'economist', 'forecast', 'gdp', 'inflation', 'fed', 'interest rate', 'recession']):
            if len(sent) > 20:

Unbounded Output

Medium
Category
Output Handling
Confidence
85% confidence
Finding

Prediction-market-related sentences are appended without length limits, again allowing untrusted remote content to control message size. The danger here is availability and output integrity rather than confidentiality: attackers can force noisy, broken, or oversized posts.

Content

Scanner excerpt · scripts/fetch_cnbc_geopolitics.py (reported line 302)May include surrounding context.

python
impacts.append(f"Analyst: {sent}")
                break
    
    # Find Polymarket/prediction market sentences - NO length limit
    for sent in sentences:
        if any(x in sent.lower() for x in ['polymarket', 'prediction', 'odds', 'probability', 'betting', 'market pricing']):
            if len(sent) > 20:

Unbounded Output

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The formatter uses complete market-impact strings and extracted facts 'as-is' from remote content, with no effective per-field truncation before assembling the outbound Discord message. Although there is a 2000-character check later, the split path is buggy and does not safely constrain arbitrary overlong fields, making output-flooding and reliability failures realistic.

Content

Scanner excerpt · scripts/fetch_cnbc_geopolitics.py (reported line 351)May include surrounding context.

python
# Market: use complete market impact string
    market = article.get('market_impact', 'See article')
    
    # Facts: use complete sentences as-is (NO truncation)
    facts = article.get('hard_facts', ['No facts extracted'])[:5]
    facts_text = ''
    for f in facts:

Unbounded Output

Medium
Category
Output Handling
Confidence
91% confidence
Finding

This loop concatenates up to five unbounded fact strings into a single message body, allowing attacker-controlled remote text to dominate output size and break delivery. In this skill, the source is external HTML and the sink is a webhook, so unbounded concatenation directly affects availability and operational safety.

Content

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

python
facts = article.get('hard_facts', ['No facts extracted'])[:5]
    facts_text = ''
    for f in facts:
        # No truncation - use complete sentence as extracted
        facts_text += f"  - {f}\n"
    
    # NEW FORMAT: Remove '### Article' and keep everything else

Unbounded Output

Medium
Category
Output Handling
Confidence
60% confidence
Finding

Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.

Content

Scanner excerpt · scripts/fetch_cnbc_geopolitics.py (reported line 115)May include surrounding context.

python
# New format template - copy and paste the whole function below

def format_single_article(article):
    """Format a SINGLE article for Discord posting (one at a time) - NO TRUNCATION."""
    if not article.get('title'):
        return None

Static analysis

No suspicious patterns detected.