Grades are a hygiene check, not a security guarantee.

Claude Code, Skill

Graph Skill Runtime

Graph Skill Runtime is a provider-neutral Python runtime for compiling, predicting, and running document-driven graph skills.

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.

PERMISSION_DELETE_FILES SECRETS_PATH_DOTENV OBFUSCATION_FILE_OVER_READ_CAP OBFUSCATION_FILE_CAP_REACHED

Why this grade

  • PERMISSION_DELETE_FILES Low

    Contains a call that can delete files.

    scripts/accept_release_artifacts.py, line 567 result = subprocess.run(

  • SECRETS_PATH_DOTENV Low

    Mentions a .env file.

    docs/mvp0/MVP0-PROGRESS-2026-05-21.md, line 152 2. **没做真实 LLM e2e** - 你后续补的 ".env 里 api key 可以用" 解了 key 依赖, 但 P0-1 还卡 Q-R-P0-1 design 拍板. 按你指示**最后一次性跑全部 e2e** (避免多次单跑烧 budget).

  • OBFUSCATION_FILE_OVER_READ_CAP Note

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

    uv.lock, line 1 version = 1

  • OBFUSCATION_FILE_CAP_REACHED Note

    The 400 file cap was reached before every file was read.

    uv.lock, line 1 version = 1

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

    scripts/accept_release_artifacts.py, line 567

    result = subprocess.run(

    scripts/accept_release_artifacts.py, line 596

    result = subprocess.run(

    scripts/validate_round28_manifest.py, line 134

    result = subprocess.run(

    Show more

    src/graph_skill_runtime/adapters/process.py, line 45

    completed = subprocess.run(argv, stdin=subprocess.DEVNULL, shell=False)

    src/graph_skill_runtime/adapters/process.py, line 47

    completed = subprocess.run(

    src/graph_skill_runtime/adapters/process.py, line 61

    process: subprocess.Popen[str]

    src/graph_skill_runtime/adapters/process.py, line 93

    subprocess.run(

    src/graph_skill_runtime/adapters/process.py, line 124

    completed = subprocess.run(

    src/graph_skill_runtime/io/storage.py, line 367

    shutil.rmtree(stale)

    src/graph_skill_runtime/migration/studio_v030.py, line 1508

    shutil.rmtree(staging_parent)

  • Install and update

    Pass

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

  • Network

    Pass

    No network call found.

  • Secrets

    Concern

    reads a credential or environment file.

    SECRETS_PATH_DOTENV

    docs/mvp0/MVP0-PROGRESS-2026-05-21.md, line 152

    2. **没做真实 LLM e2e** - 你后续补的 ".env 里 api key 可以用" 解了 key 依赖, 但 P0-1 还卡 Q-R-P0-1 design 拍板. 按你指示**最后一次性跑全部 e2e** (避免多次单跑烧 budget).

    docs/mvp0/MVP0-PROGRESS-2026-05-21.md, line 156

    ### .env API key 覆盖现状 (后续 e2e 用)

    docs/mvp0/MVP0-PROGRESS-2026-05-21.md, line 158

    | Provider 代号 | 期望 env | .env 是否有 |

    Show more

    docs/mvp0/MVP0-PROGRESS-2026-05-21.md, line 167

    `conftest.py:_load_dotenv_for_smoke()` 自动加载 .env, 不需要手 export. `premium`/`balanced`/`analyst` 角色首选 OneChats 不可用, fallback 到 WS_LLM 等; `fast`/`drafter` 角色首选 DS/ARK 直接命中.

    docs/mvp0/MVP0-PROGRESS-2026-05-21.md, line 176

    4. 我推全实施 (Block 1→4 顺序), 完成后**一次性跑全部真实 LLM e2e** (.env key 已可用, 你指示"最后一次性跑")

    docs/mvp0/MVP0-PROGRESS-2026-05-21.md, line 212

    工程实施基本卡在你拍板 18 个 Q 这一步, 醒来 30 分钟内拍完 → 我可以一夜推完 Block 1-4 全实施 + 单元/集成测试 + PR, 最后一次性跑全部真实 LLM e2e (.env key 已可用).

  • 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. the file cap was reached before every file was read.

    OBFUSCATION_FILE_OVER_READ_CAP OBFUSCATION_FILE_CAP_REACHED

    uv.lock, line 1

    version = 1

    tests/core/test_cognitive_tools_mounting.py, line 1

    """Cognitive tool mounting on the live create_agent path.

  • Licence

    Pass

    A licence file is present (LICENSE).

  • Maintainer activity

    Pass

    The latest commit is within the last 365 days.

Get the full report on this tool, £49

Note

Hygiene check of commit 205b34d34b93. can delete files. reads a credential or environment file. a file was over the read cap and was not fully read. the file cap was reached before every file was read.

Date and source

Date checked:

View source

github.com

Commit graded: 205b34d34b9354deae5fe1f095f2ba62180a9c52

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