Back to skill

Security audit

S2 SSSU Origin Alignment Brain (S2 空间原点对齐与孪生大脑)

Security checks for vulnerabilities and agentic risk

Overview

This skill is broadly a robotics/navigation control package, but it asks agents to adopt global behavior rules and includes unsafe authorization, peer-verification, and navigation-success logic that need careful review before use.

Install only in an isolated test environment, not in a real robot, smart-home, or fleet-control workflow. Treat the package as research/demo code until the host-agent directives are removed, the entry point is made executable, navigation fails closed, visas and peer signatures are cryptographically scoped, synthetic sensors are labeled as simulation-only, network permissions are narrowed, and dependency versions are locked.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • Tool Hijacking and SpoofingModifies or replaces tools so legitimate-looking calls execute attacker logic
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
Findings (10)

T01 · Skill Instruction Hijacking

Error
Location
README.md:33
Finding

Host Agent Behavior and Memory Directive Injection

Content
View full analysis
Remediation
View remediation

T05 · Unauthorized Access and Privilege Escalation

Error
Location
core/robot_navigation_pipeline.py:72
Finding

Effective Navigation Pipeline Bypasses Authorization and Physical Safety Checks

Content
View full analysis
Remediation
View remediation

T05 · Unauthorized Access and Privilege Escalation

Error
Location
core/visa_manager.py:8
Finding

Unauthenticated Visa Issuance and Missing Robot and Grid Scope Enforcement

Content
View full analysis
dict: raw_str = f"{robot_id}-{task_token}-{time.time()}" visa_token = ( f"VISA-{hashlib.md5(raw_str.encode()).hexdigest()[:8].upper()}" ) self.active_visas[visa_token] = { "robot_id": robot_id, "task_token": task_token, "allowed_grids": allowed_grids, "expires_at": time.time() + expire_seconds, "status": "VALID" } return { "status": "VISA_GRANTED", "visa_token": visa_token, "allowed_grids": allowed_grids } def validate_visa(self, visa_token: str) -> bool: if visa_token not in self.active_visas: return False if time.time() > self.active_visas[visa_token]["expires_at"]: self.active_visas[visa_token]["status"] = "EXPIRED" return False return self.active_visas[visa_token]["status"] == "VALID" ``` The governance layer only calls: ```python if not self.visa_manager.validate_visa(visa_token): return { "status": "DENIED", "reason": "Invalid or expired Spatio-Temporal Visa." } ``` ### Technical Analysis The visa-issuance action accepts caller-controlled robot identity, task token, and allowed-grid list without authenticating the caller or validating the requested scope. The validation function checks only whether the token exists, is unexpired, and has a valid status. It does not verify that: - The presented `robot_id` matches the token's recorded robot. - The requested grid is in `allowed_grids`. - The requested operation matches the authorized task. - The requester possesses an authenticated robot identity. The token co ...[truncated 1081 chars]
Remediation
View remediation

T07 · Tool Hijacking and Spoofing

Error
Location
plugins/swarm_sync.py:14
Finding

Peer Authentication Uses a Public Constant Instead of Cryptographic Verification

Content
View full analysis
bool: signature = peer_state.get("cryptographic_signature", "") if not self.fleet_pki_root: logging.error( "Fatal error: missing PKI root; network signature verification rejected." ) return False return signature == "VALID_S2_FLEET_SIG" ``` The accepted peer state is then used to calculate right-of-way: ```python peer_danger_score = self._calculate_kinematic_danger(peer_state) if my_danger_score >= peer_danger_score: decision = "PROCEED" else: decision = "YIELD" ``` ### Technical Analysis The configured PKI root is checked only for non-emptiness and is never used to verify a digital signature. Authentication succeeds whenever the caller provides the hard-coded string `VALID_S2_FLEET_SIG`. Because the expected string is visible in the source code, it is not a secret and provides no authenticity, integrity, freshness, or replay protection. The attacker also controls the peer kinematics and task data used by the arbitration calculation. ### Attack Path 1. Ensure the process has any non-empty `S2_SWARM_PKI_ROOT` value. 2. Submit a peer object with: - `cryptographic_signature` set to `VALID_S2_FLEET_SIG`. - An attacker-selected agent ID. - Attacker-selected task type, mass, and velocity. 3. `_verify_peer_signature` returns `True`. 4. The forged values are used to compute peer danger and right-of-way. 5. The attacker can induce a `YIELD` or influence other arbitration outcomes. ### Impact Assessment An unauthenticated party able to supply peer state can impersonate a fleet member and manipulate right-of-way decisions. In the first pipeline implementation, this could force a navigation halt. The ...[truncated 178 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
plugins/boundary_scanner.py:17
Finding

Boundary Scanner Misrepresents Synthetic Data as Live Hardware Input

Content
View full analysis
np.ndarray: return np.array([ [0.8, 0.0, 25.5], [0.85, 5.0, 24.0], [0.82, -5.0, 26.1], [1.5, -90.0, 5.0], [1.55, -85.0, 4.8], [5.0, 90.0, 1.0], [4.8, 180.0, 0.5] ]) def execute_boundary_scan( self, center_hex_code: str, heading_vector: str, step_size_mm: int ) -> dict: logging.info( "Reading high-frequency low-level radar point cloud and projecting topology." ) pcd_tensor = self._simulate_raw_point_cloud() ``` The effective navigation pipeline never calls `execute_boundary_scan` and directly returns success after generating state. ### Technical Analysis The scanner logs that it is reading low-level radar data, but it always consumes a fixed synthetic array. It has no hardware interface, freshness check, source authentication, or explicit simulation-mode guard. This would be unsafe even if the scanner were used because callers could mistake deterministic sample data for live physical sensing. The duplicate pipeline defect makes the situation worse by bypassing the scanner entirely. ### Attack Path 1. Deploy the Skill in a workflow that expects live boundary detection. 2. Invoke navigation with arbitrary or missing sensor data. 3. The effective pipeline does not call the scanner. 4. Even direct scanner calls return fixed synthetic points rather than live measurements. 5. The system can report a successful step without validating the actual physical boundary. ### Impact Assessment The direct software impact is false safety reporting and reliance on fabricated sensor data. If an external controller uses `STEP_SUCCESS` to authorize physi ...[truncated 193 chars]
Remediation
View remediation

T05 · Unauthorized Access and Privilege Escalation

Warning
Location
openclaw.plugin.json:11
Finding

Plugin Requests Unused Private-Network Access

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
requirements.txt:1
Finding

Automatic Installation Uses a Mutable Unpinned Dependency

Content
View full analysis
=1.24.0 ``` ```json "scripts": { "install": "pip3 install -r requirements.txt", "start": "python3 handler.py" } ``` The plugin manifest independently declares: ```json "install": "pip3 install -r requirements.txt" ``` ### Technical Analysis The npm installation lifecycle automatically invokes pip. The dependency declaration permits any NumPy version from 1.24.0 onward and supplies no package hash. This is not evidence that NumPy is malicious. The issue is that the effective package selected during a future installation can differ from the reviewed package, making builds non-reproducible and increasing supply-chain exposure. An npm lifecycle hook also crosses package ecosystems automatically during installation. ### Attack Path 1. Install the npm package or plugin. 2. The installation hook runs `pip3 install -r requirements.txt`. 3. Pip resolves the newest compatible NumPy version available from its configured index. 4. The installed artifact can differ from the version reviewed or tested by the project. 5. A compromised index, compromised future release, or incompatible version could execute or introduce unintended code in the environment. ### Impact Assessment A compromised dependency would execute with the privileges of the package installer or runtime environment. In a normal isolated installation, this is limited to the current user or container; a privileged installation could have broader system impact. No malicious third-party package was identified in the reviewed declaration. ]]>
Remediation
View remediation

other

Warning
Location
core/grid_alignment_engine.py:12
Finding

Alignment API Discloses a Hard-Coded Precise Geographic Anchor

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
handler.py:1
Finding

Declared Python Entry Point Is Not Valid Executable Python

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
core/spatial_ledger.py:6
Finding

Ledger Is Mutable Despite Being Presented as Tamper-Resistant

Content
View full analysis
str: record = { "timestamp": time.time(), "sssu_id": sssu_id, "actor": actor_id, "event_type": event_type, "payload": payload } record_string = json.dumps(record, sort_keys=True).encode() tx_hash = hashlib.sha256(record_string).hexdigest() entry = { "tx_hash": tx_hash, "record": record } self.chain.append(entry) return tx_hash def dump_ledger(self): return json.dumps(self.chain, indent=2, ensure_ascii=False) ``` ### Technical Analysis The ledger is a mutable in-memory list. Each record has an unkeyed SHA-256 hash, but: - Records are not linked to previous records. - The list can be modified or truncated. - There is no verification routine. - There is no protected signing key or message-authentication code. - There is no persistent append-only storage. - A party capable of changing a record can recompute its hash. Consequently, the implementation does not provide the tamper-resistance implied by its description. ### Attack Path 1. Obtain code-level access to the `SpatialLedger` object. 2. Modify or delete an entry in `ledger.chain`. 3. Re-serialize the modified record and recompute its SHA-256 value. 4. Replace the stored hash with the recomputed value. 5. No chain verification or external integrity control detects the modification. ### Impact Assessment An attacker with process-level object access can alter, remove, reorder, or fabricate audit events. This can conceal unauthorized calls or creat ...[truncated 121 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
Findings (63)

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

PKI verification, peer coordination, right-of-way arbitration, and environment-variable access indicate security-sensitive and distributed-control functionality beyond simple alignment. When these capabilities are hidden behind a benign-sounding description, they can evade proper review and expose secrets or affect fleet behavior.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

PKI verification, peer coordination, right-of-way arbitration, and environment-variable access indicate security-sensitive and distributed-control functionality beyond simple alignment. When these capabilities are hidden behind a benign-sounding description, they can evade proper review and expose secrets or affect fleet behavior.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

PKI verification, peer coordination, right-of-way arbitration, and environment-variable access indicate security-sensitive and distributed-control functionality beyond simple alignment. When these capabilities are hidden behind a benign-sounding description, they can evade proper review and expose secrets or affect fleet behavior.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

PKI verification, peer coordination, right-of-way arbitration, and environment-variable access indicate security-sensitive and distributed-control functionality beyond simple alignment. When these capabilities are hidden behind a benign-sounding description, they can evade proper review and expose secrets or affect fleet behavior.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

PKI verification, peer coordination, right-of-way arbitration, and environment-variable access indicate security-sensitive and distributed-control functionality beyond simple alignment. When these capabilities are hidden behind a benign-sounding description, they can evade proper review and expose secrets or affect fleet behavior.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

PKI verification, peer coordination, right-of-way arbitration, and environment-variable access indicate security-sensitive and distributed-control functionality beyond simple alignment. When these capabilities are hidden behind a benign-sounding description, they can evade proper review and expose secrets or affect fleet behavior.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

PKI verification, peer coordination, right-of-way arbitration, and environment-variable access indicate security-sensitive and distributed-control functionality beyond simple alignment. When these capabilities are hidden behind a benign-sounding description, they can evade proper review and expose secrets or affect fleet behavior.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

PKI verification, peer coordination, right-of-way arbitration, and environment-variable access indicate security-sensitive and distributed-control functionality beyond simple alignment. When these capabilities are hidden behind a benign-sounding description, they can evade proper review and expose secrets or affect fleet behavior.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

PKI verification, peer coordination, right-of-way arbitration, and environment-variable access indicate security-sensitive and distributed-control functionality beyond simple alignment. When these capabilities are hidden behind a benign-sounding description, they can evade proper review and expose secrets or affect fleet behavior.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

PKI verification, peer coordination, right-of-way arbitration, and environment-variable access indicate security-sensitive and distributed-control functionality beyond simple alignment. When these capabilities are hidden behind a benign-sounding description, they can evade proper review and expose secrets or affect fleet behavior.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

This module materially diverges from the declared skill purpose of Z-axis reduction and mandatory 2D entrance-anchor translation, instead implementing unrelated state mutation, pseudo-physics, and Monte Carlo search. In an agent skill, that scope mismatch is dangerous because it can conceal undeclared behavior, consume resources unexpectedly, and create a misleading trust boundary for operators who believe the code is alignment-only.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

The file implements broader authority than the skill’s declared purpose of spatial origin alignment, including visa validation, persistent event logging, occupant-aware decisioning, and safety isolation. In an agent system, this kind of scope expansion is dangerous because it creates hidden control-plane capabilities that can deny actions, collect sensitive operational data, or interfere with other components without being justified by the manifest.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The emergency override path can revoke visas and place a robot into an isolation state even though those powers are unrelated to origin alignment. If this skill is invoked by higher-level orchestration that assumes alignment-only behavior, an attacker or misconfigured workflow could abuse it as an undocumented kill-switch or denial-of-service mechanism against robots.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The implementation performs full robot navigation, collision handling, external negotiation, and persistent ledger writes, which substantially exceeds the declared origin-alignment scope. This kind of scope expansion is dangerous because it grants a narrowly described skill operational control and side effects that users or orchestrators may not expect, undermining least-privilege and increasing the chance of unauthorized movement or external actions.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

Injecting a generative sandbox for spatial-state synthesis is context-inappropriate for an origin-alignment skill and materially increases system power. A generative engine can fabricate or mutate perceived environment state, which in robotics could influence motion decisions and create safety hazards if downstream logic trusts the generated output.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The added generative spatial-state rendering introduces new state synthesis behavior unrelated to the stated origin-alignment purpose. In a robotics context, embedding undeclared world-state generation can mislead downstream components, create unsafe decisions, and conceal high-impact behavior inside a skill that appears limited in scope.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

This file implements issuance, validation, and revocation of access-like 'visa' tokens, which is materially broader than the stated purpose of spatial origin alignment. Even if intended for robot coordination, embedding authorization state in an alignment skill expands the attack surface and creates a hidden trust-control path that could be abused or relied upon without proper security review.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The manifest describes a skill focused on spatial twin origin alignment, Z-axis reduction, and 2D grid translation via a main entrance anchor. This file instead implements sensor fusion, collision probability estimation, and transparent-obstacle detection using camera/lidar/tactile tensors, which is a substantially different operational purpose rather than an obvious implementation detail of coordinate alignment.

Content

No source excerpt is available for this finding.

Lp1

High
Category
MCP Least Privilege
Confidence
94% confidence
Finding

The skill reads S2_SWARM_PKI_ROOT from the environment even though its declared purpose is spatial origin alignment, creating undeclared access to runtime configuration and potentially sensitive trust material. In an agent/plugin setting, undeclared capabilities are dangerous because they bypass the expected permission model and may let a mismatched or repurposed skill inspect or depend on secrets it should not access.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The implementation does not perform the advertised Z-axis reduction or 2D entrance-anchor translation; instead it performs peer signature checking and swarm right-of-way arbitration. This functionality mismatch is a strong supply-chain red flag because it can conceal unauthorized operational behavior behind an innocuous description, defeating review and permission expectations.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill performs right-of-way and kinetic hazard arbitration, including decisions to PROCEED or YIELD, which is operational control logic unrelated to origin alignment. In this context, hidden motion-priority logic is more dangerous because it can influence physical robot behavior under the cover of a benign spatial utility, potentially causing unsafe movement or unauthorized coordination.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

This line states in natural language that all important updates for the model and protocol will be recorded in Chinese. That represents a language policy constraint without offering a user choice or documenting a justified regional limitation, which matches the locale-policy violation criteria.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
81% confidence
Finding

The file explicitly labels the Chinese section as the '中文版 (Chinese Version)', indicating a language-specific presentation alongside English but without any stated user choice or explanation of locale requirements. Under the policy rule, language constraints should be optional or clearly justified; this file presents a fixed bilingual legal text with a mandated Chinese variant but no opt-in guidance.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The README explicitly instructs agents to transmit detailed robot sensor and kinematics data, but provides no data minimization guidance, consent expectations, retention limits, or privacy/security controls. In a robotics and smart-home context, such telemetry can reveal occupancy patterns, room layout, device capabilities, and movement behavior, making the omission security-relevant rather than merely informational.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill declares a required sensitive environment variable and instructs the agent to invoke a tool that may consume it, but there is no user-facing disclosure or consent boundary. Hidden secret usage is dangerous because operators may unknowingly trigger privileged trust-root access or downstream authenticated operations.

Content

No source excerpt is available for this finding.