Grades are a hygiene check, not a security guarantee.

MCP, MCP server

io.github.SaravananJaichandar/world-model-mcp

Persistent memory + optional signed audit for AI coding agents.

Compatibility: Not stated.

Grade B

Minor issues only, no Medium or High. Worth a quick look at the findings below before installing.

Graded by PlugVetRubric v2

Not a guarantee. A hygiene check of the source at the commit shown, not a security audit.

Rescanned at commit 5e14e4b71203, unchanged since the check below.

PERMISSION_DELETE_FILES NETWORK_NO_FIXED_HOST SECRETS_PATH_DOTENV SECRETS_ENV_READ OBFUSCATION_FILE_OVER_READ_CAP

Why this grade

  • PERMISSION_DELETE_FILES Low

    Contains a call that can delete files.

    benchmarks/repeat-mistake/agent_runner.py, line 184 proc = subprocess.run(

  • NETWORK_NO_FIXED_HOST Low

    Makes a network call whose host is not written as a fixed address.

    benchmarks/repeat-mistake/select_subset.py, line 84 " curl -sL -o benchmarks/repeat-mistake/verified.parquet "

  • SECRETS_PATH_DOTENV Low

    Mentions a .env file.

    world_model_server/mcp_tool_dictionary.py, line 47 of a file matching `.env` / `credentials`). That's a v2 concern;

  • SECRETS_ENV_READ Note

    Reads an environment variable whose name contains KEY, SECRET, TOKEN, PASSWORD or CREDENTIAL.

    benchmarks/coach-player/run.py, line 103 api_key = os.environ.get("ANTHROPIC_API_KEY")

  • OBFUSCATION_FILE_OVER_READ_CAP Note

    A file was over the read cap and was not fully read.

    benchmarks/repeat-mistake/baseline_results_s2.jsonl, line 1 {"instance_id": "matplotlib__matplotlib-14623", "resolved": true, "tests_status": {"FAIL_TO_PASS": {"success": ["lib/matplotlib/tests/test_axes.py::test_inverted_limits"], "failure

Any High, or two or more Medium findings, give D. One Medium gives C. Low findings or a security Note give B. Each code names a finding behind the letter.

Grades are a hygiene check of the source at the commit shown, on the date checked, not a security audit or a guarantee. The free directory grades are automated. Before release, a PlugVet engineer reviews every paid report and any grade we hold back, including all F grades, against the source code and the evidence we quote.

What was checked

Plain notes from the hygiene checks.

  • Permissions

    Concern

    can delete files.

    PERMISSION_DELETE_FILES

    benchmarks/repeat-mistake/agent_runner.py, line 184

    proc = subprocess.run(

    benchmarks/repeat-mistake/claude_client.py, line 86

    proc = subprocess.run(

    benchmarks/repeat-mistake/claude_client.py, line 135

    proc = subprocess.run(

    Show more

    benchmarks/repeat-mistake/clone_repo.py, line 39

    proc = subprocess.run(

    benchmarks/repeat-mistake/clone_repo.py, line 145

    shutil.rmtree(checkout.cleanup_token, ignore_errors=True)

    benchmarks/repeat-mistake/score.py, line 56

    proc = subprocess.run(cmd, env=os.environ.copy())

    examples/managed-agents-self-hosted/deploy_modal.py, line 54

    subprocess.Popen(["python", "-m", "world_model_server.server"])

    hooks/src/world-model-session.ts, line 88

    fs.unlinkSync(eventsQueue);

    scripts/engagement_tracker.py, line 78

    result = subprocess.run(

    world_model_server/cli.py, line 505

    result = subprocess.run(

    world_model_server/hooks/world-model-session.js, line 102

    fs.unlinkSync(eventsQueue);

  • Install and update

    Pass

    No install-time script found, and dependencies are pinned or absent.

  • Network

    Concern

    can reach a host that is not written as a fixed address.

    NETWORK_NO_FIXED_HOST

    benchmarks/repeat-mistake/select_subset.py, line 84

    " curl -sL -o benchmarks/repeat-mistake/verified.parquet "

    world_model_server/mcp_tool_dictionary.py, line 95

    # string should upgrade to `critical` on sudo, rm -rf, curl to

    world_model_server/telemetry.py, line 307

    with urllib.request.urlopen(req, timeout=4) as resp:

    Show more

    world_model_server/telemetry.py, line 419

    with urllib.request.urlopen(req, timeout=4) as resp:

    benchmarks/repeat-mistake/treatment_progress_seed2_treatment.jsonl, line 6

    {"instance_id": "sympy__sympy-17630", "arm": "treatment", "success": false, "extracted_patch": "diff --git a/sympy/matrices/expressions/matexpr.py b/sympy/matrices/expressions/mate

  • Secrets

    Concern

    reads a credential or environment file.

    SECRETS_ENV_READ SECRETS_PATH_DOTENV

    benchmarks/coach-player/run.py, line 103

    api_key = os.environ.get("ANTHROPIC_API_KEY")

    world_model_server/config.py, line 21

    default_factory=lambda: os.getenv("ANTHROPIC_API_KEY")

    world_model_server/mcp_tool_dictionary.py, line 47

    of a file matching `.env` / `credentials`). That's a v2 concern;

    Show more

    world_model_server/config.py, line 67

    default_factory=lambda: os.getenv("WORLD_MODEL_VERIFICATION_API_KEY"),

    world_model_server/seeder.py, line 28

    "venv", ".venv", "env", ".env", ".tox", "coverage", ".mypy_cache",

    world_model_server/tools.py, line 82

    or os.getenv("OPENROUTER_API_KEY")

    world_model_server/tools.py, line 83

    or os.getenv("OPENAI_API_KEY")

    .env.example, line 2

    # Copy this file to .env and fill in your values

    .env.example, line 3

    # IMPORTANT: Never commit .env files to git!

    .env.example, line 64

    # 1. This file should be copied to .env in your project root

    .env.example, line 65

    # 2. Never commit .env files - they contain sensitive credentials

    .env.example, line 66

    # 3. The .gitignore file automatically excludes .env files

    .gitignore, line 60

    # Environment

  • Agent instructions

    Pass

    No instruction to ignore earlier instructions or to hide a step.

  • Obfuscation

    Note

    a file was over the read cap and was not fully read.

    OBFUSCATION_FILE_OVER_READ_CAP

    benchmarks/repeat-mistake/baseline_results_s2.jsonl, line 1

    {"instance_id": "matplotlib__matplotlib-14623", "resolved": true, "tests_status": {"FAIL_TO_PASS": {"success": ["lib/matplotlib/tests/test_axes.py::test_inverted_limits"], "failure

    benchmarks/repeat-mistake/treatment_results_s2_crossdomain.jsonl, line 1

    {"instance_id": "matplotlib__matplotlib-14623", "resolved": true, "tests_status": {"FAIL_TO_PASS": {"success": ["lib/matplotlib/tests/test_axes.py::test_inverted_limits"], "failure

    benchmarks/repeat-mistake/verified.parquet, line 1

    PAR1��L�,astropy/

  • Licence

    Pass

    A licence file is present (LICENSE).

  • Maintainer activity

    Pass

    The latest commit is within the last 365 days.

Note

Hygiene check of commit 5e14e4b71203. can delete files. can reach a host that is not written as a fixed address. reads a credential or environment file. a file was over the read cap and was not fully read.

Date and source

Date checked:

View source

github.com

Commit graded: 5e14e4b71203c993cb2d38ad068ebef6fa5da02f

This link goes to the original host. PlugVet does not host or check what you download from there, and installing it is your own decision and risk.

Read how this letter is chosen