Back to skill

Security audit

Seedream 5.0 Lite 精准文字图片

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent AI Hive image-generation wrapper with disclosed API-key setup, uploads, task polling, and downloads, but users should treat its text-accuracy checks as manual and note a few implementation hardening gaps.

Install only if you are comfortable giving this skill an AI Hive API key and letting it upload selected reference images and download generated outputs. Do not rely on it to automatically verify exact text; manually compare any generated text before publishing, especially for prices, dates, product claims, legal terms, or regulated content.

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 (2)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/imagegen.py:501
Finding

API Key File Is Created Before Restrictive Permissions Are Applied

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Note
Location
scripts/imagegen.py:263
Finding

Remote Generation Results Are Downloaded Without Destination or Size Restrictions

Content
View full analysis
0: pct = int(downloaded * 100 / total) bar = "█" * (pct // 5) + "░" * (20 - pct // 5) print(f"\r {bar} {pct}% ({downloaded // 1024}KB)", end="", flush=True) print() size_mb = downloaded / (1024 * 1024) print(f"[ok] {out_path} ({size_mb:.1f} MB)") except requests.exceptions.RequestException as e: print(f"\n[error] Download failed: {e}", file=sys.stderr) ``` The function is invoked for API-provided result URLs at `scripts/imagegen.py:358` and last-frame URLs at `scripts/imagegen.py:365`. ### Technical Analysis Task-response fields such as `resultUrl` and `lastFrameUrl` are passed directly to `requests.get()`. The implementation does not enforce: - An HTTPS-only URL scheme. - An allowlist of approved object-storage hosts. - Restrictions on cross-host redirects. - A maximum response size. - A maximum streamed byte count. - An expected image content type or file signature. - Cleanup of partial files following an interrupted or oversized download. Streaming prevents the entire response from being loaded into memory, but it does not prevent disk exhaustion. A missing or falsified `Co ...[truncated 1533 chars]
Remediation
View remediation
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (6)

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

The skill description claims exact-text verification, character budgeting, and fallback controls, but the described implementation does not actually perform OCR, text comparison, or automated validation and also includes extra operational abilities not clearly disclosed. This is dangerous because users may rely on non-existent safety and accuracy guarantees for commercial text images, leading to publication of incorrect regulated content, while the hidden management capabilities broaden the attack surface beyond the declared purpose.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
82% confidence
Finding

The skill exposes capabilities implying environment access, file read/write, network, and shell execution, but does not declare an explicit tool scope or permission boundary. This is dangerous because consumers of the skill cannot clearly understand or constrain what the skill may access, increasing the risk of unintended file modification, secret exposure, or outbound network use if the underlying scripts are invoked in a broader agent environment.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The module docstring and subsequent user-facing CLI messages are written entirely in Chinese, and the file does not indicate that users may choose another language or that the skill is intentionally limited to a Chinese-only region/compliance context. Under the stated policy, a forced language/locale without user opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The description claims capabilities such as literal proofreading, character-budget control, and failure fallback to improve text accuracy. In practice, the code only fixes the model ID, forwards prompt/images/params to the backend, and polls/downloads results; it does not inspect text, enforce budgets, validate OCR output, or implement retry/fallback logic for text fidelity.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The example API key value uses Chinese-only instructional text ('请替换为你的APIKey'), which imposes a specific language in user-facing content. For a general configuration example, this can violate language/locale policy when no alternative locale or opt-in is provided.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Low
Category
Not specified by scanner
Confidence
78% confidence
Finding

The manifest frames this skill as generating or editing Seedream 5 Lite commercial images with text-accuracy controls. While API authentication is expected, actively launching a browser and guiding interactive account/API-key setup is an auxiliary account-management capability not justified by the stated image-generation purpose.

Content

No source excerpt is available for this finding.