Back to skill

Security audit

volcengine-tts-feishu

Security checks for vulnerabilities and agentic risk

Overview

The skill does what it claims, but its Feishu sending path uses local app credentials and has concrete credential-handling and temporary-file safety weaknesses.

Review before installing if you plan to use Feishu sending. Use narrowly scoped Feishu app permissions, avoid sensitive text unless you accept third-party processing by Volcengine and Feishu, and prefer a fixed version of the dependency manifest with requests declared. The publisher should replace curl with an in-process HTTP upload, avoid putting tokens in command-line arguments, and create temporary files atomically.

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 (3)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/http_tts.py:153
Finding

Predictable temporary audio file permits symlink-based file overwrite

Content
View full analysis
.mp3` as a symbolic link to another file writable by the victim. 3. The victim invokes the Skill with `--send-to`. 4. The script opens the predictable path with write-and-truncate behavior. 5. The linked target is overwritten with synthesized audio. 6. Alternatively, the attacker races to replace or modify the temporary audio before `ffmpeg` reads it, potentially changing the content sent to Feishu. Successful exploitation requires local access, accurate ...[truncated 690 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/http_tts.py:27
Finding

Feishu bearer token is exposed through curl process arguments

Content
View full analysis
` in its command-line arguments. 4. A local attacker monitors process metadata through an available process-listing or accounting interface. 5. The attacker captures the bearer token before the `curl` process exits. 6. The attacker submits Feishu API requests with that token until it expires or is invalidated. Exploitability depends on the operating system's process visibility rules and the attacker's local access. ### Impact Assessment A captured token may permit unauthorized Feishu API operations within the scopes granted to the configured application. Depending on those scopes, consequences may include unauthorized file uploads, messaging ...[truncated 370 chars]
Remediation
View remediation

T08 · Insecure Dependencies

Note
Location
requirements.txt:1
Finding

Dependency manifest is unbounded and omits the required HTTP library

Content
View full analysis
=12.0 ``` The implementation instead imports an undeclared package: ```python import requests ``` ### Technical Analysis The dependency manifest permits any current or future version of `websockets` at or above version 12.0, even though the reviewed implementation does not use that package. This unnecessarily expands the installed third-party code and prevents reproducible dependency resolution. Conversely, `requests`, which is directly imported and used for the TTS request, is not declared. A clean installation following the documented `pip install -r requirements.txt` procedure may therefore fail at runtime or rely on an ambient, unaudited version already installed in the environment. This is a dependency-integrity weakness rather than evidence that the named package is currently malicious. The risk arises from unbounded future resolution, unnecessary package installation, and reliance on undeclared environment state. ### Attack Path 1. A user creates a new environment and runs `pip install -r requirements.txt`. 2. The package installer resolves the newest available `websockets` release satisfying the unbounded constraint. 3. Unnecessary third-party code is installed despite not being used by the Skill. 4. The required `requests` package is not installed by the manifest. 5. Execution either fails at import time or succeeds only because an ambient version of `requests` happens to be present. 6. The environment consequently becomes non-reproducible and may consume future or locally supplied package versions that were not reviewed with the Skill. Exploitation would require compromise or manipulation of the selected package source, dependency resolution environment, or ambien ...[truncated 636 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (17)

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 19)May include surrounding context.

json
{
  "appid": "你的AppID",
  "access_token": "你的Access Token",
  "default_resource_id": "seed-tts-2.0",
  "default_voice": "zh_female_meilinvyou_uranus_bigtts",
  "default_emotion": "vocal-fry"

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 45)May include surrounding context.

json
{
  "appid": "你的AppID",
  "access_token": "你的Access Token",
  "default_resource_id": "seed-tts-2.0",
  "default_voice": "zh_female_meilinvyou_uranus_bigtts",
  "default_emotion": "vocal-fry"

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 67)May include surrounding context.

json
{
  "appid": "你的AppID",
  "access_token": "你的Access Token",
  "default_resource_id": "seed-tts-2.0",
  "default_voice": "zh_female_meilinvyou_uranus_bigtts",
  "default_emotion": "vocal-fry"

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 105)May include surrounding context.

json
{
  "appid": "你的AppID",
  "access_token": "你的Access Token",
  "default_resource_id": "seed-tts-2.0",
  "default_voice": "zh_female_meilinvyou_uranus_bigtts",
  "default_emotion": "vocal-fry"

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 119)May include surrounding context.

json
{
  "appid": "你的AppID",
  "access_token": "你的Access Token",
  "default_resource_id": "seed-tts-2.0",
  "default_voice": "zh_female_meilinvyou_uranus_bigtts",
  "default_emotion": "vocal-fry"

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/http_tts.py (reported line 61)May include surrounding context.

python
```json
{
  "appid": "你的AppID",
  "access_token": "你的Access Token",
  "default_resource_id": "seed-tts-2.0",
  "default_voice": "zh_female_meilinvyou_uranus_bigtts",
  "default_emotion": "vocal-fry"

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
91% confidence
Finding

The skill documentation describes capabilities that read local files, perform network operations, and invoke shell commands, but it does not declare any tool scope or permission boundaries. This creates an authorization and transparency gap: users and orchestration systems cannot easily restrict the skill to the minimum required privileges before use.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill encourages use of --send-to to transmit synthesized audio to Feishu but does not clearly warn that both message content-derived audio and recipient identifiers are sent over the network to third-party services. This can lead to unintended disclosure of sensitive or personal data, especially if users synthesize confidential text.

Content

No source excerpt is available for this finding.

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
70% confidence
Finding

Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Content

Scanner excerpt · SKILL.md (reported line 199)May include surrounding context.

md
brew install ffmpeg

# Ubuntu/Debian
sudo apt install ffmpeg

# CentOS
sudo yum install ffmpeg

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
70% confidence
Finding

Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Content

Scanner excerpt · SKILL.md (reported line 202)May include surrounding context.

sudo apt install ffmpeg

CentOS

sudo yum install ffmpeg

text

如果没有ffmpeg,`--send-to` 功能无法使用。只生成MP3不需要ffmpeg。

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/http_tts.py (reported line 29)May include surrounding context.

python
return json.loads(r.read())["tenant_access_token"]

def upload_opus_file(token, opus_path):
    result = subprocess.run([
        "curl", "-s", "-X", "POST",
        f"{FEISHU_API}/im/v1/files",
        "-H", f"Authorization: Bearer {token}",

External Transmission

Medium
Category
Data Exfiltration
Confidence
80% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · scripts/http_tts.py (reported line 130)May include surrounding context.

python
payload["req_params"]["emotion"] = args.emotion
    
    print(f"Synthesizing: {args.text}")
    response = requests.post(url, headers=headers, json=payload, stream=True)
    
    if response.status_code != 200:
        print(f"Error {response.status_code}: {response.text}")

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

When --send-to is used, the script transmits synthesized audio and the recipient open_id to Feishu APIs without a strong, explicit disclosure at execution time. In a skill ecosystem, silent external transfer of generated content and identifiers increases privacy and data-exfiltration risk, especially if users do not realize local text is being converted and sent to third-party services.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

The script silently reads Feishu app credentials from a local OpenClaw config and immediately uses them to obtain an access token and perform authenticated actions. In an agent-skill context, that can cause unintended use of locally available secrets and outbound actions under the user's identity or tenant without explicit confirmation.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/http_tts.py (reported line 187)May include surrounding context.

python
with tempfile.TemporaryDirectory() as tmpdir:
            opus_path = os.path.join(tmpdir, "voice.opus")
            # ffmpeg转Opus
            result = subprocess.run(
                ["ffmpeg", "-i", output_path, "-c:a", "libopus", opus_path, "-y"],
                capture_output=True
            )

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
95% confidence
Finding

The dependency is specified as websockets>=12.0, which permits any future major or minor release and makes builds non-reproducible. This increases supply-chain risk because a later incompatible or vulnerable version could be installed without review, and it also prevents reliable verification against known advisories.

Content

Scanner excerpt · requirements.txt (reported line 1)May include surrounding context.

text
websockets>=12.0

Unverifiable Dependency: websockets has 4 known advisory(ies) (CVE-2018-1000518 (websockets is vulnerable to denial of service by memory exhaustion); CVE-2021-33880 (Observable Timing Discrepancy in aaugustin websockets library); CVE-2018-1000518 (aaugustin websockets version 4 contains a CWE-409: Improper Handling of Highly C) +1 more), but the manifest does not pin a version, so it is unknown whether the installed release is affected

Low
Category
Supply Chain
Confidence
88% confidence
Finding

Because websockets is not pinned, it is impossible to determine from this manifest whether deployment will use a version affected by known advisories. In a network-facing TTS/streaming skill, a vulnerable websocket library could expose the service to denial-of-service or other library-specific weaknesses if that code path is exercised.

Content

No source excerpt is available for this finding.