Back to skill

Security audit

Sano Intel · 探针资本

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent healthcare intelligence API wrapper, but it sends the API token and user queries over unencrypted HTTP.

Install only if you are comfortable sending your Sano API token and query contents to this service over plaintext HTTP. Prefer an HTTPS endpoint with valid certificate verification before using real tokens, and avoid storing long-lived tokens in shell startup files when possible.

Vulnerability Patterns
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (1)

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:56
Finding

API Credential Transmitted over Unencrypted HTTP

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 56–58. The same insecure transport pattern is used by every documented API request. Vulnerability Type: Plaintext transmission of an API credential Risk Level: High

bash
curl -s "http://47.102.196.1:8081/v1/sector_heat?top_n=10" \
  -H "X-API-Key: $SANO_TOKEN" \
  -H "Content-Type: application/json"

Technical Analysis

The skill instructs the agent to transmit SANO_TOKEN in the X-API-Key header to an endpoint using plaintext HTTP. HTTP provides neither transport encryption nor server authentication. Consequently, any party capable of observing or modifying network traffic can read the API credential, inspect potentially sensitive queries and responses, or alter the returned healthcare and financial intelligence.

This pattern affects all documented API operations, including company searches, financing events, clinical trials, market information, daily digests, and patent queries. Using a raw IP address also prevents meaningful hostname-based identity validation if TLS is later introduced without a properly validated certificate and domain.

Attack Path

  1. A user obtains a valid Sano API token and exports it as SANO_TOKEN.
  2. The agent follows SKILL.md and sends a request to http://47.102.196.1:8081.
  3. An attacker with an on-path position, such as a malicious Wi-Fi operator, compromised router, network administrator, or upstream intermediary, captures the plaintext HTTP request.
  4. The attacker extracts the token from the X-API-Key header.
  5. The attacker reuses the token to issue unauthorized API requests within the permissions and usage limits assigned to the victim.
  6. Alternatively, the attacker modifies API responses in transit, causing the agent to present manipulated company, financing, clinical-trial, patent, or market information as legitimate data.

Impact Assessment

Successful exploitation exposes the victim's AP ...[truncated 557 chars]

Remediation
View remediation

Remediation Suggestions

  1. Replace every http://47.102.196.1:8081 endpoint with an authenticated https:// endpoint hosted under a controlled domain.
  2. Configure the server with a certificate issued for that domain by a trusted certificate authority.
  3. Preserve curl's default certificate and hostname verification. Do not introduce --insecure, -k, or custom logic that bypasses TLS verification.
  4. Disable or reject plaintext HTTP access instead of relying solely on redirects, because the credential may already be exposed before a redirect is processed.
  5. Rotate all API tokens that may previously have been transmitted through the plaintext endpoint.
  6. Use short-lived, revocable, least-privilege tokens and apply server-side rate limits and anomaly detection.
  7. Avoid persisting tokens directly in shell startup files where practical. Prefer an operating-system credential store, secret manager, or protected configuration file with restrictive permissions.
  8. Update every example in SKILL.md and verify through automated checks that no credential-bearing request can use a non-TLS URL.
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (3)

External Transmission

Medium
Category
Data Exfiltration
Confidence
99% confidence
Finding

The skill explicitly requires transmitting the API key in an X-API-Key header to an external service over plaintext HTTP. Because this skill handles proprietary market, company, and research queries, both the credential and the user's request contents could be intercepted, replayed, or tampered with during transit.

Content

Scanner excerpt · SKILL.md (reported line 23)May include surrounding context.

md
## 数据查询方式

所有查询通过 curl 调用 Sano Intel API,**必须带上** `X-API-Key: $SANO_TOKEN` header。

API Base:`http://47.102.196.1:8081`

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill instructs users to send an API token and query data to an HTTP endpoint without transport encryption. This exposes credentials and potentially sensitive user queries to interception or modification by any on-path attacker, and the documentation does not warn about that risk.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The usage principles state that when users ask in Chinese the skill should automatically select an API and that results should be summarized in Chinese. This imposes a language behavior without documenting a user choice or opt-in, which can violate language/locale policy requirements.

Content

No source excerpt is available for this finding.