Security audit

Price Watcher

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent price watcher, but it can persistently fetch any stored URL and optionally send alert contents to Discord without enough scoping or warning.

Install only if you are comfortable with a local script storing product URLs and price history, fetching those URLs repeatedly, and optionally sending formatted alerts to Discord. Avoid adding internal, localhost, metadata-service, credential-bearing, or untrusted URLs, and review webhook output before enabling scheduled notifications.

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 (3)

T05 · Unauthorized Access and Privilege Escalation

Error

Location: scripts/add_product.py:96
Finding: Unrestricted URL Fetching Enables Server-Side Request Forgery
Content: View full analysis
= 3: try: manual_price = float(sys.argv[2]) except ValueError: print(f"[ERROR] Price must be a number, got: {sys.argv[2]}") sys.exit(1) if len(sys.argv) >= 4: manual_name = sys.argv[3] price = manual_price name = manual_name if price is None or name is None: print(f"[INFO] Fetching page: {url}") html, err = fetch_page(url) ``` ```python # scripts/check_prices.py:225-234 for i, product in enumerate(products): url = product["url"] name = product["name"] old_price = product.get("last_price") or product.get("baseline_price") if i > 0: time.sleep(1) # polite delay between requests print(f"[INFO] Checking: {name}", file=sys.stderr) html, err = fetch_page(url) ``` ### Technical Analysis Both scripts pass user- ...[truncated 2495 chars]
Remediation: View remediation

T09 · Insecure Skill Coding Practices

Warning

Location: scripts/add_product.py:96
Finding: Unbounded HTTP Response Read Can Exhaust Process Memory
Content: View full analysis
Remediation: View remediation
MAX_RESPONSE_SIZE: return None, "Response exceeds maximum allowed size" ``` 3. Check `Content-Length` before reading, while retaining the runtime limit because the header may be absent or dishonest. 4. Stream the response in bounded chunks rather than allowing an unrestricted allocation. 5. Accept only expected content types such as `text/html` and `application/xhtml+xml`. 6. Apply consistent response-size, timeout, and content-type controls in both fetching scripts. 7. Consider process-level memory and CPU limits for scheduled or agent-triggered executions. ]]>

T09 · Insecure Skill Coding Practices

Warning

Location: scripts/format_alert.py:51
Finding: Untrusted Product Metadata Is Rendered as Raw Markdown
Content: View full analysis
tag.""" m = re.search(r']*>(.*?)', html, re.IGNORECASE | re.DOTALL) if m: title = m.group(1).strip() # Strip site suffixes like "| Amazon" or "- Best Buy" title = re.split(r'\s*[|–—-]\s*(?:Amazon|eBay|Walmart|Best Buy|Target|Costco|Newegg)', title)[0] return title[:120].strip() return None ``` ```python # scripts/add_product.py:172 name = extract_title_from_html(html) or url ``` ```python # scripts/format_alert.py:51-82 for item in sorted(drops, key=lambda x: x.get("change_pct", 0)): lines.append(f"### {item['name']}") lines.append(f"- **Was:** {format_price(item.get('old_price'))}") lines.append(f"- **Now:** {format_price(item.get('new_price'))}") lines.append(f"- **Change:** {format_change(item.get('change_amount'), item.get('change_pct'))}") lines.append(f"- **Link:** [{item['url']}]({item['url']})") lines.append("") if increases: lines.append("## 📈 Price Increases") lines.append("") for item in sorted(increases, key=lambda x: x.get("change_pct", 0), reverse=True): lines.append(f"### {item['name']}") lines.append(f"- **Was:** {format_price(item.get('old_price'))}") lines.append(f"- **Now:** {format_price(item.get('new_price'))}") lines.append(f"- **Change:** {format_change(item.get('change_amount'), item.get('change_pct'))}") lines.append(f"- **Link:** [{item['url']}]({item['url']})") lines.append("") if errors: lines.append("## ⚠️ Check Errors") lines.append("") for item in errors: status = item.get("status", "error") err_msg = item.get("e ...[truncated 2330 chars]
Remediation: View remediation

SkillSpector

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (7)

External Script Fetching

High

Category: Supply Chain
Confidence: 90% confidence
Finding: Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.
Content: Scanner excerpt · references/setup-guide.md (reported line 45)May include surrounding context.
python3 scripts/format_alert.py >> /tmp/price-alerts.md 2>/tmp/price-watcher.log

text
Or to send results to Discord via curl (replace WEBHOOK_URL): ```cron 0 9 * * * cd /Users/openclaw/.openclaw/workspace/skills/price-watcher && \

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The trigger phrases are broad enough to match common shopping or product-related requests, which can cause the skill to activate when the user did not explicitly consent to starting persistent monitoring. In this skill, activation can lead to local state changes and outbound fetching of arbitrary product URLs, so overbroad routing increases the chance of unintended file modification and network access.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The guide includes a cron example that posts generated output to a Discord webhook but does not clearly warn users that this sends skill-produced data to an external third-party service. Even if the payload appears limited to price-watch results, product names, URLs, and other derived content may be transmitted off-host, creating an avoidable data disclosure risk through documentation-driven configuration.
Content: No source excerpt is available for this finding.

External Transmission

Medium

Category: Data Exfiltration
Confidence: 88% confidence
Finding: This command sends the formatted results to the URL stored in DISCORD_WEBHOOK, which is an external network transmission of locally generated data. In context this is intentional functionality rather than covert exfiltration, but it still creates a real outbound data path that could disclose tracked products, pricing history, or injected content from watched pages to an external service.
Content: Scanner excerpt · references/setup-guide.md (reported line 50)May include surrounding context.
cron
0 9 * * * cd /Users/openclaw/.openclaw/workspace/skills/price-watcher && \ RESULT=$(python3 scripts/check_prices.py | python3 scripts/compare.py | python3 scripts/format_alert.py) && \ [ -n "$RESULT" ] && curl -s -X POST "$DISCORD_WEBHOOK" \ -H "Content-Type: application/json" \ -d "{\"content\": \"\`\`\`\n$RESULT\n\`\`\`\"}"

Missing User Warnings

Low

Category: Not specified by scanner
Confidence: 87% confidence
Finding: The skill explicitly states that it updates watchlist.json, but it does not prominently warn the user that running the workflow will modify local data. This can lead to surprising persistence of tracked URLs, product names, and pricing history, especially when the user may expect a read-only check.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The request headers hard-code "Accept-Language: en-US,en;q=0.9", which imposes a specific language/locale preference regardless of user choice. This is a natural-language policy concern because the skill does not offer any opt-in, override, or justification for forcing English.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The HTTP request hard-codes the Accept-Language header to "en-US,en;q=0.9", which imposes a specific language/locale preference. This is a natural-language policy concern because the skill does not offer user opt-in or explain why English is required.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.