Back to skill

Security audit

Micro-Doppler Life-Safety Engine

Security checks for vulnerabilities and agentic risk

Overview

The skill is disclosed as smart-home eldercare automation, but real mode can unlock a door and send a Home Assistant bearer token to a configurable HTTP endpoint with too little safety control.

Keep this skill in dry-run unless you have reviewed the device actions and network path. Before enabling real actuation, remove or gate door unlocking, require human approval or independent sensor confirmation, use HTTPS with a fixed trusted Home Assistant host, use a least-privilege token, and pin dependencies with a lockfile.

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

Error
Location
skill.py:16
Finding

Home Assistant bearer token can be transmitted over plaintext HTTP

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
skill.py:39
Finding

Safety-critical physical actions lack authenticated sensor input and runtime authorization

Content
View full analysis
10.0 ``` A positive event automatically triggers physical actions, including unlocking a door: ```python if event_type == "FALL_DETECTED": HomeAssistantActuator.call_service( "lock", "unlock", "lock.room_802_main_door" ) HomeAssistantActuator.call_service( "light", "turn_on", "light.room_802_all", {"brightness_pct": 100, "color_temp": 200} ) HomeAssistantActuator.call_service( "fan", "set_percentage", "fan.room_802_hvac", {"percentage": 100} ) ``` ### Technical Analysis The single environment variable `S2_ENABLE_REAL_ACTUATION` changes the system from simulation to unrestricted automatic physical actuation. Once enabled, the router does not require user confirmation, a second independent sensor, event authenticity, per-action authorization, or a policy check before unlocking the configured door. The input pipeline does not currently consume authenticated hardware data. It constructs a predefined synthetic fall profile with random noise. Conseq ...[truncated 1788 chars]
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
requirements.txt:1
Finding

Unbounded and unhashed third-party dependencies permit unreproducible installations

Content
View full analysis
=1.24.0 scipy>=1.10.0 matplotlib>=3.7.0 requests>=2.28.0 ``` The installation instructions execute dependency resolution directly: ```bash pip install -r requirements.txt ``` ### Technical Analysis Every dependency uses an open-ended minimum-version constraint. A future installation can therefore resolve to versions that were not present during this audit. The project also supplies no lockfile and no cryptographic hashes for package artifacts. The listed package names are established projects, and the reviewed files contain no evidence of typosquatting or a malicious package. The issue is that the installation process does not preserve the reviewed dependency set or verify exact artifacts. A compromised future release, unsafe transitive dependency, incompatible update, or package-index compromise could change the code installed and executed in the skill environment. ### Attack Path 1. A user runs `pip install -r requirements.txt`. 2. The package resolver selects the newest versions satisfying the open-ended constraints. 3. A selected direct or transitive package version differs from the set previously reviewed and tested. 4. If that package release or distribution artifact is compromised, its installation or runtime code executes in the skill environment. 5. The malicious code receives the privileges of the account installing or running the skill and may access the same environment, including configured Home Assistant credentials. This is a supply-chain exposure rather than evidence that any currently named dependency is malicious. ### Impact Assessment Potential impact depends on the privileges of the installation and runtime account. Compromised dependency code could read environment variables, modify files accessible to ...[truncated 177 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
Findings (13)

Context-Inappropriate Capability

Critical
Category
Not specified by scanner
Confidence
98% confidence
Finding

This code can issue live unlock, lighting, and HVAC commands to external Home Assistant services, including unlocking a door during a fall-detection path. In context, that is especially dangerous because the trigger logic appears weak or incorrect, so a false positive or manipulated configuration could produce unauthorized physical actions affecting occupant safety and property security.

Content

No source excerpt is available for this finding.

Tainted flow: 'url' from os.getenv (line 68, credential/environment) → requests.post (network output)

Critical
Category
Data Flow
Confidence
95% confidence
Finding

The request target is built from an environment-controlled base URL and then used in a live POST with an Authorization bearer token. If real actuation is enabled, a malicious or misconfigured HA_BASE_URL can redirect commands and credentials to an attacker-controlled endpoint, enabling SSRF-like outbound access and token exfiltration alongside physical-device command abuse.

Content

Scanner excerpt · skill.py (reported line 81)May include surrounding context.

python
# 真实物理控制逻辑
        print(f"      └─ 🔌 [物理执行层] 正在发送危险指令 POST {url} | 载荷: {payload}")
        try:
            response = requests.post(url, headers=HomeAssistantActuator.headers, json=payload, timeout=5)
            response.raise_for_status()
            print(f"      └─ ✅ [硬件响应] 成功执行真实物理指令!")
            return True

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 · manifest.json (reported line 12)May include surrounding context.

json
"environment_variables": {
    "S2_ENABLE_REAL_ACTUATION": "Default: False. Set to 'True' to allow actual HTTP requests to physical hardware.",
    "HA_BASE_URL": "The REST API endpoint for Home Assistant (e.g., http://localhost:8123/api).",
    "HA_BEARER_TOKEN": "Long-lived access token for Home Assistant authentication."
  },
  "language": "python",
  "homepage": "https://space2.world"

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill is explicitly wired for smart-home actuation through environment-provided Home Assistant connectivity and a runtime flag that can enable real device control. Even though dry-run is the default, the presence of production actuation capability in a skill with no authentication, authorization, safety policy, or user-confirmation layer creates meaningful physical-security risk.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The file includes prominent natural-language descriptions and runtime messaging in Chinese, but nowhere indicates that language is configurable or user-selected. Under the policy, forcing a specific language without opt-in is a natural-language policy violation unless the locale constraint is clearly documented and justified.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
90% confidence
Finding

The skill performs external network transmission of command payloads and authorization headers to a configurable endpoint. External transmission alone is not always harmful, but here it carries control commands and a bearer token for physical automation, making confidentiality and destination integrity materially important.

Content

Scanner excerpt · skill.py (reported line 81)May include surrounding context.

python
# 真实物理控制逻辑
        print(f"      └─ 🔌 [物理执行层] 正在发送危险指令 POST {url} | 载荷: {payload}")
        try:
            response = requests.post(url, headers=HomeAssistantActuator.headers, json=payload, timeout=5)
            response.raise_for_status()
            print(f"      └─ ✅ [硬件响应] 成功执行真实物理指令!")
            return True

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
96% confidence
Finding

The dependency is specified with a lower bound only, which allows future installs to resolve to different versions over time. This weakens reproducibility and can unintentionally introduce vulnerable or incompatible releases through the software supply chain.

Content

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

text
numpy>=1.24.0
scipy>=1.10.0
matplotlib>=3.7.0
requests>=2.28.0

Unverifiable Dependency: numpy has 16 known advisory(ies) (CVE-2014-1859 (Numpy arbitrary file write via symlink attack); CVE-2021-41495 (NumPy NULL Pointer Dereference); CVE-2021-33430 (NumPy Buffer Overflow (Disputed)) +13 more), but the manifest does not pin a version, so it is unknown whether the installed release is affected

Low
Category
Supply Chain
Confidence
40% confidence
Finding

Dependency has known vulnerabilities (CVEs). Using packages with unpatched security flaws exposes the environment to known exploits.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
96% confidence
Finding

The dependency is not pinned to an exact version, so builds are non-deterministic and may pull in an unsafe or untested scipy release later. This creates a supply-chain risk even if no immediately exploitable issue is evident in the file itself.

Content

Scanner excerpt · requirements.txt (reported line 2)May include surrounding context.

text
numpy>=1.24.0
scipy>=1.10.0
matplotlib>=3.7.0
requests>=2.28.0

Unverifiable Dependency: scipy has 4 known advisory(ies) (CVE-2013-4251 (SciPy creates insecure temporary directories); CVE-2013-4251 (The scipy.weave component in SciPy before 0.12.1 creates insecure temporary dire); CVE-2023-25399 (A refcounting issue which leads to potential memory leak was discovered in scipy) +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
40% confidence
Finding

Dependency has known vulnerabilities (CVEs). Using packages with unpatched security flaws exposes the environment to known exploits.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
95% confidence
Finding

Using a minimum-only version for matplotlib permits uncontrolled upgrades during installation. That can lead to unexpected behavior or exposure to future vulnerable versions, making the environment less predictable and less secure.

Content

Scanner excerpt · requirements.txt (reported line 3)May include surrounding context.

text
numpy>=1.24.0
scipy>=1.10.0
matplotlib>=3.7.0
requests>=2.28.0

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
98% confidence
Finding

An unpinned requests dependency is more concerning because it is commonly used for network communication and has a history of security advisories. Allowing arbitrary newer releases at install time can expose the skill to transport, credential-handling, or request-processing flaws.

Content

Scanner excerpt · requirements.txt (reported line 4)May include surrounding context.

text
numpy>=1.24.0
scipy>=1.10.0
matplotlib>=3.7.0
requests>=2.28.0

Unverifiable Dependency: requests has 16 known advisory(ies) (CVE-2014-1830 (Exposure of Sensitive Information to an Unauthorized Actor in Requests); CVE-2024-47081 (Requests vulnerable to .netrc credentials leak via malicious URLs); CVE-2024-35195 (Requests `Session` object does not verify requests after making first request wi) +13 more), but the manifest does not pin a version, so it is unknown whether the installed release is affected

Low
Category
Supply Chain
Confidence
40% confidence
Finding

Dependency has known vulnerabilities (CVEs). Using packages with unpatched security flaws exposes the environment to known exploits.

Content

No source excerpt is available for this finding.