T09 · Insecure Skill Coding Practices
- Location
SKILL.md:6- Finding
Unencrypted Communication with an External Financial Data API
- Content
View full analysis
Vulnerability Details
File Location:
SKILL.md, line 6
Vulnerability Type: Plaintext HTTP transport to an external API
Risk Level: MediumVulnerable Code Snippet:
text HTTP HOST: http://60.205.179.76:8000Technical Analysis
The Skill instructs the agent to obtain stock information, financial indicators, shareholder data, and corporate announcements from an external API using plaintext HTTP. HTTP provides neither transport encryption nor server authentication.
An attacker with a network-level interception capability can observe requests and modify responses in transit. Because the endpoint is identified by a bare IP address and does not use TLS, the client cannot verify that returned market data, announcement metadata, PDF paths, or extracted announcement content originated from the intended service.
The exposed material does not contain evidence of credentials or authentication tokens. Nevertheless, stock queries may disclose user research interests, while response manipulation may compromise the integrity of subsequent financial analysis. This finding does not establish that the remote server itself is malicious or vulnerable.
Attack Path
- An agent loads the Skill and constructs a request using
http://60.205.179.76:8000. - The request traverses a network controlled or monitored by an attacker, such as an untrusted wireless network, compromised gateway, malicious proxy, or hostile upstream provider.
- The attacker intercepts the plaintext HTTP request and observes the requested company, date range, search terms, or announcement keywords.
- The attacker modifies the API response to return fabricated prices, financial indicators, shareholder figures, announcement metadata, or announcement text.
- The agent treats the modified response as legitimate API data and may use it in research summaries or downstream decisions.
Impact Assessment
No operating-system privileges, loc ...[truncated 678 chars]
- An agent loads the Skill and constructs a request using
- Remediation
View remediation
Remediation Suggestions
- Replace the plaintext endpoint with an HTTPS endpoint protected by a valid, publicly trusted TLS certificate.
- Use a verifiable domain name associated with the service operator rather than an undocumented bare IP address.
- Enable strict TLS certificate and hostname validation; do not permit insecure fallback to HTTP.
- Document the service owner, data-handling policy, availability expectations, and trust boundary.
- Validate all API responses against strict schemas, including expected data types, date formats, stock-code formats, size limits, and permitted path formats.
- Treat all returned financial data, announcement text, and file-path values as untrusted external input.
- Where data integrity is important, corroborate critical market or announcement information against an authoritative independent source.
- Avoid placing credentials, tokens, personal information, or other sensitive material in query strings.