Back to skill

Security audit

Multi-Agent Cross-Device Encrypted Collaboration (Email + Task Sealing)

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because the scanners report a mismatch between its stated collaboration purpose and supplied examples, along with an unsafe handling concern for external summaries; the overall confidence is medium.

Consider the suspicious verdict and incomplete analysis before installing. In particular, weigh the external-summary execution concern and the mismatch between the promised collaboration features and the behavior described by the scanner; also consider whether the reported file and network capabilities fit your needs.

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
examples/role_prompts.md
Finding

External result summaries are promoted to executable code

Content
View full analysis

Impact

The workflow gives an external executor a code-execution entry point in the orchestrator's execution environment rather than limiting it to summary submission. Depending on that environment's permissions, malicious report logic could access local files or modify orchestrator artifacts.

Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • 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 (9)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The supplied code's sole behavior is printing a static template. Although a task template could support the declared workflow, this placeholder does not implement the skill's stated primary functionality. No undeclared resource access or malicious behavior is evident.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The code has a materially different apparent purpose from the declared skill: it illustrates importing a shared helper. The comments suggest a module-path error, but the literal ellipsis in the second import causes a syntax error before execution. The imported helper's implementation is not supplied, so no additional behavior can be inferred.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
80% confidence
Finding

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

Content

Scanner excerpt · SKILL.md (reported line 344)May include surrounding context.

bash
# Local debug
python3 examples/use_credential.py --local --token-file ./api_token.txt

# Sealed run (called by executor agent inside mgc_run)
# ext02='["--owner","local_credentials","--name","api_token.txt"]'

Credential Access

High
Category
Privilege Escalation
Confidence
80% confidence
Finding

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

Content

Scanner excerpt · SKILL.md (reported line 347)May include surrounding context.

python3 examples/use_credential.py --local --token-file ./api_token.txt

Sealed run (called by executor agent inside mgc_run)

ext02='["--owner","local_credentials","--name","api_token.txt"]'

text

---

Credential Access

High
Category
Privilege Escalation
Confidence
80% confidence
Finding

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

Content

Scanner excerpt · examples/use_credential.py (reported line 31)May include surrounding context.

python
info_type="script",
         info_owner="<info_owner inside package>",
         diff_3="use_credential.py",
         ext02='["--owner","local_credentials","--name","api_token.txt"]',
     )

  3) Local debug (plaintext, reads a disk file directly):

Credential Access

High
Category
Privilege Escalation
Confidence
80% confidence
Finding

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

Content

Scanner excerpt · examples/use_credential.py (reported line 35)May include surrounding context.

python
)

  3) Local debug (plaintext, reads a disk file directly):
     python3 use_credential.py --local --token-file ./api_token.txt

Zero third-party dependencies (stdlib urllib).
"""

Credential Access

High
Category
Privilege Escalation
Confidence
80% confidence
Finding

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

Content

Scanner excerpt · examples/use_credential.py (reported line 22)May include surrounding context.

python
info_type="text",
         info_owner="local_credentials",
         diff_2="credential",
         diff_3="api_token.txt",
         content="sk-your-token",
     )

Credential Access

High
Category
Privilege Escalation
Confidence
80% confidence
Finding

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

Content

Scanner excerpt · examples/use_credential.py (reported line 113)May include surrounding context.

python
help="info_owner of the credential in local MGC")
    parser.add_argument("--diff-2", default="credential",
                        help="diff_2 of the credential in local MGC")
    parser.add_argument("--name", default="api_token.txt",
                        help="diff_3 of the credential in local MGC")
    parser.add_argument("--token-file", default=None,
                        help="Local debug only: read credential directly from this disk file")

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: file_read, file_write, network.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.