Trust Assessment
vitonique/a2a-secure:root received a trust score of 10/100, placing it in the Untrusted category. This skill has significant security findings that require attention before use in production.
SkillShield's automated analysis identified 75 findings: 26 critical, 38 high, 11 medium, and 0 low severity. Key findings include Network egress to untrusted endpoints, Unsafe environment variable passthrough, Arbitrary command execution.
The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. The Manifest Analysis layer scored lowest at 0/100, indicating areas for improvement.
Last analyzed on August 1, 2026 (commit d3da7071). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings75
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | send.py:69 | |
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | send.py:70 | |
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | send.py:425 | |
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | tests/test_admin_api.py:84 | |
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | tests/test_admin_api.py:333 | |
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | tests/test_audit_api.py:47 | |
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | tests/test_google_a2a.py:1353 | |
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | tests/test_google_a2a.py:1363 | |
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | tests/test_send_hardening.py:99 | |
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | tests/test_send_hardening.py:103 | |
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | tests/test_send_hardening.py:113 | |
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | tests/test_send_hardening.py:119 | |
| CRITICAL | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | tools/send_delegated_ping.py:63 | |
| CRITICAL | Network egress to untrusted endpoints Python requests POST/PUT to URL Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | tools/test_delegation_shim.py:62 | |
| CRITICAL | Network egress to untrusted endpoints Python requests POST/PUT to URL Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | tools/test_ecrecover.py:60 | |
| CRITICAL | Arbitrary command execution Python shell execution (os.system, subprocess) Review all shell execution calls. Ensure commands are static (not built from user input), use absolute paths, and are strictly necessary. Prefer library APIs over shell commands. | Manifest | send.py:399 | |
| CRITICAL | Arbitrary command execution Python shell execution (os.system, subprocess) Review all shell execution calls. Ensure commands are static (not built from user input), use absolute paths, and are strictly necessary. Prefer library APIs over shell commands. | Manifest | send.py:402 | |
| CRITICAL | Arbitrary command execution Python shell execution (os.system, subprocess) Review all shell execution calls. Ensure commands are static (not built from user input), use absolute paths, and are strictly necessary. Prefer library APIs over shell commands. | Manifest | tools/intro_approval.py:65 | |
| CRITICAL | Arbitrary command execution Python shell execution (os.system, subprocess) Review all shell execution calls. Ensure commands are static (not built from user input), use absolute paths, and are strictly necessary. Prefer library APIs over shell commands. | Manifest | tools/local_harness.py:202 | |
| CRITICAL | Arbitrary command execution Python shell execution (os.system, subprocess) Review all shell execution calls. Ensure commands are static (not built from user input), use absolute paths, and are strictly necessary. Prefer library APIs over shell commands. | Manifest | tools/node_lifecycle.py:42 | |
| CRITICAL | Arbitrary command execution Python shell execution (os.system, subprocess) Review all shell execution calls. Ensure commands are static (not built from user input), use absolute paths, and are strictly necessary. Prefer library APIs over shell commands. | Manifest | tools/node_provision.py:68 | |
| CRITICAL | Arbitrary command execution Python shell execution (os.system, subprocess) Review all shell execution calls. Ensure commands are static (not built from user input), use absolute paths, and are strictly necessary. Prefer library APIs over shell commands. | Manifest | tools/trustreg.py:280 | |
| CRITICAL | Arbitrary command execution Python shell execution (os.system, subprocess) Review all shell execution calls. Ensure commands are static (not built from user input), use absolute paths, and are strictly necessary. Prefer library APIs over shell commands. | Manifest | wake.py:42 | |
| CRITICAL | Credential harvesting Bulk environment variable dump Skills should only access environment variables they explicitly need. Bulk environment dumps (os.environ.copy, JSON.stringify(process.env)) are almost always malicious. Remove access to Keychain, GPG keys, and credential stores. | Manifest | tests/test_trust.py:622 | |
| CRITICAL | Credential harvesting Bulk environment variable dump Skills should only access environment variables they explicitly need. Bulk environment dumps (os.environ.copy, JSON.stringify(process.env)) are almost always malicious. Remove access to Keychain, GPG keys, and credential stores. | Manifest | tools/intro_approval.py:61 | |
| CRITICAL | Credential harvesting Bulk environment variable dump Skills should only access environment variables they explicitly need. Bulk environment dumps (os.environ.copy, JSON.stringify(process.env)) are almost always malicious. Remove access to Keychain, GPG keys, and credential stores. | Manifest | tools/node_spec.py:80 | |
| HIGH | Unsafe environment variable passthrough Bulk environment variable harvesting Minimize environment variable exposure. Only pass required, non-sensitive variables to MCP servers. Use dedicated secret management instead of environment passthrough. | Manifest | tests/test_trust.py:622 | |
| HIGH | Unsafe environment variable passthrough Bulk environment variable harvesting Minimize environment variable exposure. Only pass required, non-sensitive variables to MCP servers. Use dedicated secret management instead of environment passthrough. | Manifest | tools/intro_approval.py:61 | |
| HIGH | Unsafe environment variable passthrough Bulk environment variable harvesting Minimize environment variable exposure. Only pass required, non-sensitive variables to MCP servers. Use dedicated secret management instead of environment passthrough. | Manifest | tools/node_spec.py:80 | |
| HIGH | Unsafe deserialization / dynamic eval Decryption followed by code execution Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tests/test_admin_api.py:10 | |
| HIGH | Unsafe deserialization / dynamic eval Decryption followed by code execution Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tests/test_characterization.py:31 | |
| HIGH | Unsafe deserialization / dynamic eval Decryption followed by code execution Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tests/test_keystore.py:4 | |
| HIGH | Unsafe deserialization / dynamic eval Decryption followed by code execution Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tests/test_server_lifecycle.py:196 | |
| HIGH | Unsafe deserialization / dynamic eval Decryption followed by code execution Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tools/cp_inventory.py:3 | |
| HIGH | Unsafe deserialization / dynamic eval Decryption followed by code execution Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tools/local_harness.py:131 | |
| HIGH | Unsafe deserialization / dynamic eval Decryption followed by code execution Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tools/local_harness.py:367 | |
| HIGH | Unsafe deserialization / dynamic eval Decryption followed by code execution Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tools/node.py:14 | |
| HIGH | Unsafe deserialization / dynamic eval Decryption followed by code execution Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tools/node_lifecycle.py:3 | |
| HIGH | Unsafe deserialization / dynamic eval Decryption followed by code execution Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tools/tenant_registry.py:31 | |
| HIGH | Unsafe deserialization / dynamic eval Python builtins/import manipulation Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tests/test_google_a2a.py:865 | |
| HIGH | Unsafe deserialization / dynamic eval Python builtins/import manipulation Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tests/test_identity_roundtrip.py:58 | |
| HIGH | Unsafe deserialization / dynamic eval Python builtins/import manipulation Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tests/test_identity_roundtrip.py:59 | |
| HIGH | Unsafe deserialization / dynamic eval Python builtins/import manipulation Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tests/test_identity_roundtrip.py:72 | |
| HIGH | Unsafe deserialization / dynamic eval Python builtins/import manipulation Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tests/test_identity_roundtrip.py:73 | |
| HIGH | Unsafe deserialization / dynamic eval Python builtins/import manipulation Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tests/test_identity_roundtrip.py:85 | |
| HIGH | Unsafe deserialization / dynamic eval Python builtins/import manipulation Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | tests/test_identity_roundtrip.py:86 | |
| HIGH | Unsafe deserialization / dynamic eval Python builtins/import manipulation Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions. | Manifest | wake.py:24 | |
| HIGH | Dangerous call: subprocess.check_output() Call to 'subprocess.check_output()' detected in function 'sync_notify'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | send.py:399 | |
| HIGH | Dangerous call: subprocess.check_output() Call to 'subprocess.check_output()' detected in function 'sync_notify'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | send.py:402 | |
| HIGH | Potential data exfiltration: file read + network send Function 'send_message' reads files and sends data over the network. This may indicate data exfiltration. Review this function to ensure file contents are not being sent to external servers. | Static | send.py:377 | |
| HIGH | Dangerous call: __import__() Call to '__import__()' detected in function 'test_ed25519_delegation_sign_verify_roundtrip'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | tests/test_identity_roundtrip.py:58 | |
| HIGH | Dangerous call: __import__() Call to '__import__()' detected in function 'test_ed25519_delegation_sign_verify_roundtrip'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | tests/test_identity_roundtrip.py:59 | |
| HIGH | Dangerous call: __import__() Call to '__import__()' detected in function 'test_ed25519_delegation_wrong_key_fails'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | tests/test_identity_roundtrip.py:72 | |
| HIGH | Dangerous call: __import__() Call to '__import__()' detected in function 'test_ed25519_delegation_wrong_key_fails'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | tests/test_identity_roundtrip.py:73 | |
| HIGH | Dangerous call: __import__() Call to '__import__()' detected in function 'test_ed25519_delegation_wrong_ts_fails'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | tests/test_identity_roundtrip.py:85 | |
| HIGH | Dangerous call: __import__() Call to '__import__()' detected in function 'test_ed25519_delegation_wrong_ts_fails'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | tests/test_identity_roundtrip.py:86 | |
| HIGH | Dangerous call: subprocess.call() Call to 'subprocess.call()' detected in function '_run_onboard'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | tools/intro_approval.py:65 | |
| HIGH | Dangerous call: subprocess.run() Call to 'subprocess.run()' detected in function '_signed_request'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | tools/local_harness.py:202 | |
| HIGH | Dangerous call: subprocess.Popen() Call to 'subprocess.Popen()' detected in function 'start_node'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | tools/node_lifecycle.py:42 | |
| HIGH | Dangerous call: subprocess.run() Call to 'subprocess.run()' detected in function '_generate_keys'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | tools/node_provision.py:68 | |
| HIGH | Dangerous call: subprocess.run() Call to 'subprocess.run()' detected in function 'send_wake'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | wake.py:42 | |
| HIGH | Dangerous call: __import__() Call to '__import__()' detected in function 'send_wake'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | wake.py:24 | |
| HIGH | World-readable sensitive data stored in shared temporary directory The application defaults to using the system temporary directory (/tmp on Unix/Linux) as the base state directory when state_dir is not configured. It creates directories like a2a-messages, a2a-google-tasks, and a2a-idempotency under this path using os.makedirs without restricting permissions. Consequently, these directories are created with world-readable permissions (typically 0o755), allowing any local user on a shared system to read sensitive message blobs, task data, artifacts, and cached responses. Change the default state directory to a user-specific secure directory (e.g., ~/.local/share/a2a or ~/.config/a2a), or ensure that any directories created under /tmp are initialized with restrictive permissions (e.g., 0o700) using os.makedirs(path, mode=0o700) and explicitly setting the mode if the directory already exists. | LLM | config.py:76 | |
| HIGH | Unauthenticated privileged Admin API on localhost allows CSRF and DNS Rebinding The tools/admin_api.py service binds to 127.0.0.1 by default and makes authentication optional if A2A_ADMIN_TOKEN is not set. This allows any local process, or any external website visited by a user on the same machine (via DNS Rebinding or Cross-Origin Resource Sharing/CORS-bypassing simple requests), to interact with the privileged Admin API without authentication. This can lead to unauthorized node registration, tenant/workspace manipulation, and control plane takeover. Enforce authentication by default even on localhost binds. Generate a secure random admin token on startup if one is not configured, and print it to the console or save it to a secure local file for the operator to use. | LLM | tools/admin_api.py:15 | |
| MEDIUM | Suspicious import: urllib.request Import of 'urllib.request' detected. This module provides network or low-level system access. Verify this import is necessary. Network and system modules in skill code may indicate data exfiltration. | Static | send.py:26 | |
| MEDIUM | Suspicious import: urllib.request Import of 'urllib.request' detected. This module provides network or low-level system access. Verify this import is necessary. Network and system modules in skill code may indicate data exfiltration. | Static | tests/test_admin_api.py:25 | |
| MEDIUM | Suspicious import: urllib.request Import of 'urllib.request' detected. This module provides network or low-level system access. Verify this import is necessary. Network and system modules in skill code may indicate data exfiltration. | Static | tests/test_audit_api.py:19 | |
| MEDIUM | Suspicious import: urllib.request Import of 'urllib.request' detected. This module provides network or low-level system access. Verify this import is necessary. Network and system modules in skill code may indicate data exfiltration. | Static | tools/local_harness.py:140 | |
| MEDIUM | Suspicious import: requests Import of 'requests' detected. This module provides network or low-level system access. Verify this import is necessary. Network and system modules in skill code may indicate data exfiltration. | Static | tools/send_delegated_ping.py:2 | |
| MEDIUM | Suspicious import: requests Import of 'requests' detected. This module provides network or low-level system access. Verify this import is necessary. Network and system modules in skill code may indicate data exfiltration. | Static | tools/test_delegation_shim.py:3 | |
| MEDIUM | Suspicious import: requests Import of 'requests' detected. This module provides network or low-level system access. Verify this import is necessary. Network and system modules in skill code may indicate data exfiltration. | Static | tools/test_ecrecover.py:3 | |
| MEDIUM | Unpinned Python dependency version Dependency 'cryptography>=41' is not pinned to an exact version. Pin Python dependencies with exact versions where feasible. | Dependencies | pyproject.toml | |
| MEDIUM | Unpinned Python dependency version Requirement 'cryptography>=41' is not pinned to an exact version. Pin Python dependencies with '==<exact version>'. | Dependencies | requirements.txt:2 | |
| MEDIUM | Unpinned Python dependency version Requirement 'eth-account>=0.13' is not pinned to an exact version. Pin Python dependencies with '==<exact version>'. | Dependencies | requirements.txt:3 | |
| MEDIUM | Replay protection check is vulnerable to TOCTOU race condition The check_replay function in hardening.py checks if a trace ID lock file exists using os.path.exists(cache_path) and then creates it using open(cache_path, "w"). This introduces a Time-of-Check to Time-of-Use (TOCTOU) race condition. In concurrent environments (e.g., multi-threaded servers or multiple instances sharing the same cache directory), two identical requests arriving simultaneously can both pass the existence check before either writes the lock file, effectively bypassing replay protection. Use atomic file creation to check and write the lock file in a single operation. In Python, this can be achieved by opening the file with the "x" mode (exclusive creation), which raises a FileExistsError if the file already exists. | LLM | hardening.py:59 |
Scan History
Embed Code
[](https://skillshield.io/report/dba55d498057ee4a)
Powered by SkillShield