agy-mcp

v0.1.8 suspicious
7.0
High Risk

Skill-first, MCP-second bridge from Claude Code / OpenAI Codex to Google Antigravity CLI (agy).

🤖 AI Analysis

Final verdict: SUSPICIOUS

The package exhibits significant signs of potential malicious behavior, including high risks related to credential harvesting and code obfuscation, which could be indicative of a supply-chain attack.

  • High credential risk due to attempts to access sensitive files
  • Code obfuscation techniques potentially used to evade detection
Per-check LLM notes
  • Obfuscation: The presence of regex compilation and import statements within a try-except block suggests potential for code obfuscation or evasion techniques.
  • Credentials: Patterns indicate attempts to access sensitive files such as /etc/passwd and ~/.ssh/id_rsa, which are commonly associated with credential harvesting activities.
  • Metadata: The maintainer has only one package, which may indicate a new or less active account, raising some suspicion but not conclusive evidence of malice.

📦 Package Quality Overall: Medium (5.6/10)

◈ Medium Test Suite 6.0

Partial test coverage signals detected

  • Test runner config found: pyproject.toml
◈ Medium Documentation 5.0

Some documentation present

  • Detailed PyPI description (6717 chars)
○ Low Contributing Guide 4.0

No contributing guide or governance files found

  • Development Status classifier >= Beta
◈ Medium Type Annotations 5.0

Partial type annotation coverage

  • 212 type-annotated function signatures detected in source
✦ High Multiple Contributors 8.0

Active multi-contributor project

  • 3 unique contributor(s) across 80 commits in Boulea7/agy-mcp
  • Small but multi-author team (3–4 contributors)

🔬 Heuristic Checks

✓ Outbound Network Calls

No suspicious network call patterns found

⚠ Code Obfuscation score 4.0

Found 2 obfuscation pattern(s)

  • _RE_AGENT_EXECUTOR_ERROR = re.compile(r"agent executor error: (.+)") _RE_FAILED_PRECONDITION = re.compile(r"FA
  • rpreter.""" try: __import__(name) return True except Exception: # noqa: BLE001 -
⚠ Shell / Subprocess Execution score 10.0

Found 6 shell execution pattern(s)

  • nvocation when needed. ``subprocess.Popen(argv, shell=False)`` cannot reliably execute a ``.cmd``/
  • False try: proc = subprocess.run( # noqa: S603 - argv hard-coded ["git", "rev-pa
  • .""" try: proc = subprocess.run( # noqa: S603 - argv hard-coded ["git", "rev-pa
  • ] try: proc = subprocess.run( # noqa: S603 - argv hard-coded argv,
  • th)) try: proc = subprocess.run( # noqa: S603 - argv hard-coded argv,
  • """ try: proc = subprocess.run( # noqa: S603 - argv hard-coded ["git", "branch
⚠ Credential Harvesting score 10.0

Found 5 credential access pattern(s)

  • oned flag like ``--extra-env "/etc/passwd=x"`` to reach the env. Values containing NUL / CR / LF
  • leak but a hostile key like ``/etc/passwd`` would otherwise land # raw in the envelope. Phase
  • malicious ``settings.json -> ~/.ssh/id_rsa`` cannot trick us into echoing private content into
  • y-cli/log/ (e.g. a symlink to ~/.ssh/id_rsa) could be read and its path leaked into events. """
  • h/"), re.compile(r"(?im)/\.aws/credentials"), re.compile(r"(?im)/\.config/(gcloud|gh|git/credentia
✓ Typosquatting

No typosquatting candidates detected

✓ Registered Email Domain

No author email provided

✓ Suspicious Page Links

All external links appear legitimate

✓ Git Repository History

Repository Boulea7/agy-mcp appears legitimate

⚠ Maintainer History score 2.0

1 maintainer concern(s) found

  • Author "Boulea7" appears to have only 1 package on PyPI (new or inactive account)
✓ Known CVE Vulnerabilities

No known vulnerabilities found in OSV database.

💡 AI App Starter Prompt

Use this prompt to build a project with agy-mcp 
Develop a command-line utility called 'CodeGravity' that leverages the 'agy-mcp' Python package to enable developers to seamlessly integrate Claude Code/OpenAI Codex with Google Antigravity CLI (agy). This utility will serve as a skill-first, MCP-second bridge, allowing users to write Python code snippets or scripts using natural language and have them executed via agy commands.

Key Features:
1. **Natural Language Input**: Users can describe their coding tasks in plain English, and the utility will generate the corresponding Python code using Claude Code/OpenAI Codex.
2. **Automatic Execution via agy**: Once the code is generated, it will automatically be piped into agy for execution, utilizing its powerful CLI capabilities.
3. **Interactive Mode**: An interactive mode where users can continuously input tasks and receive immediate feedback on the results of their code execution.
4. **History Log**: A feature that logs all user inputs and outputs for review and reference.
5. **Error Handling**: Robust error handling to guide users through common mistakes and provide solutions or workarounds.
6. **Custom Commands**: Allow users to define custom commands that can be executed directly through agy.

Steps to Develop:
1. **Setup Environment**: Install necessary dependencies including 'agy-mcp', Claude Code/OpenAI Codex API client, and any other required libraries.
2. **Interface Design**: Create a user-friendly interface that accepts natural language descriptions of coding tasks.
3. **Code Generation**: Implement functionality to convert these descriptions into executable Python code using Claude Code/OpenAI Codex.
4. **Execution via agy**: Integrate 'agy-mcp' to pipe the generated code into agy for execution.
5. **Feedback Mechanism**: Develop a system to provide immediate feedback to users regarding the success or failure of their code execution.
6. **History and Logging**: Ensure all interactions are logged for future reference.
7. **Testing**: Thoroughly test the utility with various scenarios to ensure reliability and accuracy.
8. **Documentation**: Provide comprehensive documentation explaining how to use the utility and its features.
9. **Deployment**: Prepare the utility for deployment, ensuring it can be easily installed and run on different systems.