Trust Assessment
tencentcloud-cos-skill 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 12 findings: 4 critical, 6 high, 2 medium, and 0 low severity. Key findings include Arbitrary command execution, Dangerous call: subprocess.Popen(), Dangerous call: subprocess.run().
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 February 13, 2026 (commit 13146e6a). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings12
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| 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 | skills/shawnminh/tencent-cloud-cos/scripts/cos_wrapper.py:109 | |
| 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 | skills/shawnminh/tencent-cloud-cos/test_skill.py:62 | |
| 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 | skills/shawnminh/tencent-cloud-cos/test_skill.py:73 | |
| 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 | skills/shawnminh/tencent-cloud-cos/test_skill.py:253 | |
| HIGH | Dangerous call: subprocess.Popen() Call to 'subprocess.Popen()' detected in function 'start_mcp_server'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | skills/shawnminh/tencent-cloud-cos/scripts/cos_wrapper.py:109 | |
| HIGH | Dangerous call: subprocess.run() Call to 'subprocess.run()' detected in function 'check_dependencies'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | skills/shawnminh/tencent-cloud-cos/test_skill.py:62 | |
| HIGH | Dangerous call: subprocess.run() Call to 'subprocess.run()' detected in function 'check_dependencies'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | skills/shawnminh/tencent-cloud-cos/test_skill.py:73 | |
| HIGH | Dangerous call: subprocess.run() Call to 'subprocess.run()' detected in function 'run_example'. This can execute arbitrary code. Avoid using dangerous functions like exec/eval/os.system. Use safer alternatives. | Static | skills/shawnminh/tencent-cloud-cos/test_skill.py:253 | |
| HIGH | Unpinned Dependency in Installation The installation instructions in `SKILL.md` and `install.sh` use `npm install -g cos-mcp@latest`. Relying on `@latest` for a dependency means that any new version, including potentially malicious or breaking changes, will be automatically installed without review. This introduces a significant supply chain risk, as a compromise of the `cos-mcp` package could directly affect the security of this skill. Pin the `cos-mcp` dependency to a specific, known-good version (e.g., `cos-mcp@1.2.3`) to ensure consistency and prevent unexpected or malicious updates. Regularly review and manually update the pinned version. | LLM | SKILL.md:61 | |
| HIGH | Credentials Exposed in Process Arguments The `_build_mcp_command` function in `scripts/cos_wrapper.py` constructs a command that includes `SecretId` and `SecretKey` directly as command-line arguments (e.g., `--SecretId=... --SecretKey=...`). When this command is executed via `subprocess.Popen` in `start_mcp_server`, these sensitive credentials can be visible in the system's process list (`ps aux`) to other users or processes on the same machine. This is a common method of credential exposure. Pass sensitive credentials to child processes via environment variables instead of command-line arguments. Environment variables are generally not visible in `ps aux` output. Alternatively, use a secure IPC mechanism or temporary files with strict permissions and guaranteed deletion if environment variables are not feasible. | LLM | scripts/cos_wrapper.py:90 | |
| MEDIUM | Sensitive environment variable access: $HOME Access to sensitive environment variable '$HOME' detected in shell context. Verify this environment variable access is necessary and the value is not exfiltrated. | Static | skills/shawnminh/tencent-cloud-cos/install.sh:153 | |
| MEDIUM | Temporary File Creation with Sensitive Credentials The `_call_mcp_tool` function in `scripts/cos_wrapper.py` includes code to create a temporary JSON file containing `SecretId` and `SecretKey` using `tempfile.NamedTemporaryFile(delete=False)`. Although the subsequent usage of this file is currently mocked, the presence of this code pattern indicates an intent to write sensitive credentials to disk. The `delete=False` flag means the file would persist unless explicitly removed. If the skill were to crash or be interrupted after creating this file but before cleanup, the credentials would remain on the filesystem, potentially accessible to unauthorized entities depending on the temporary directory's permissions. Avoid writing sensitive credentials to disk, even temporarily. If absolutely necessary, ensure temporary files are created with restrictive permissions (e.g., `mode=0o600`) and are *always* deleted immediately after use, preferably within a `try...finally` block or by using `tempfile.TemporaryFile` which handles deletion automatically. | LLM | scripts/cos_wrapper.py:160 |
Scan History
Embed Code
[](https://skillshield.io/report/a5c8c0a05d28268d)
Powered by SkillShield