Claude Code, Skill
Building With Typesafe Jev
Unofficial Claude Code skill for building with TypeSafe AI's Jev: the typed Choice/Score/Noul API, calibrated-confidence routing, field-tested anti-patterns, and links to 150+ community projects sorted by implementation shape, with a code sketch per pattern.
Compatibility: Not stated.
Grade D
Two or more Medium findings, or a High. Only try it in a sandbox.
Not a guarantee. A hygiene check of the source at the commit shown, not a security audit.
INSTALL_REMOTE_SCRIPT SECRETS_PATH_CREDENTIAL
Why this grade
- INSTALL_REMOTE_SCRIPT Medium
Fetches a script from the network and runs it (a pipe to a shell, or an install hook with a URL).
scripts/install_muse.sh, line 12
# curl -fsSL https://raw.githubusercontent.com/aaddrick/building-with-typesafe-jev/main/scripts/install_muse.sh | bash - SECRETS_PATH_CREDENTIAL Medium
Mentions a credential store path (~/.ssh, id_rsa, .npmrc, .aws/credentials, credentials.json, .netrc).
evals/container/run.sh, line 17
creds=~/.claude/.credentials.json
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.
If you think this grade is wrong, email [email protected] with the entry link and your evidence. We aim to acknowledge within two working days and decide within seven. If we confirm an error, we aim to correct it within 24 hours of confirming it. These are aims, not guarantees.
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
Note
requests shell access. The manifest names this permission.
PERMISSION_SHELL_DECLARED
evals/lib/rejudge.py, line 100
text = subprocess.run(["git", "show", f"{commit}:{rel}"], capture_output=True, text=True, check=True).stdout
Install and update
Serious
install fetches a script from the network. README documents an installer that pipes a download into a shell. The docs document an installer that pipes a download into a shell.
INSTALL_REMOTE_SCRIPT INSTALL_DOCUMENTED_INSTALLER
README.md, line 131
curl -fsSL https://raw.githubusercontent.com/aaddrick/building-with-typesafe-jev/main/scripts/install_muse.sh | bash
scripts/install_muse.sh, line 12
# curl -fsSL https://raw.githubusercontent.com/aaddrick/building-with-typesafe-jev/main/scripts/install_muse.sh | bash
.github/readme/README.it.md, line 131
curl -fsSL https://raw.githubusercontent.com/aaddrick/building-with-typesafe-jev/main/scripts/install_muse.sh | bash
Show more
.github/readme/README.ja.md, line 131
curl -fsSL https://raw.githubusercontent.com/aaddrick/building-with-typesafe-jev/main/scripts/install_muse.sh | bash
.github/readme/README.ko.md, line 131
curl -fsSL https://raw.githubusercontent.com/aaddrick/building-with-typesafe-jev/main/scripts/install_muse.sh | bash
.github/readme/README.pt-BR.md, line 131
curl -fsSL https://raw.githubusercontent.com/aaddrick/building-with-typesafe-jev/main/scripts/install_muse.sh | bash
.github/readme/README.vi.md, line 131
curl -fsSL https://raw.githubusercontent.com/aaddrick/building-with-typesafe-jev/main/scripts/install_muse.sh | bash
.github/readme/README.zh-CN.md, line 131
curl -fsSL https://raw.githubusercontent.com/aaddrick/building-with-typesafe-jev/main/scripts/install_muse.sh | bash
.github/workflows/plugin-load-check.yml, line 64
curl -fsSL https://antigravity.google/cli/install.sh | bash
Network
Concern
can reach a host that is not written as a fixed address. can reach codeload.github.com. The description does not name that host.
NETWORK_NO_FIXED_HOST NETWORK_UNNAMED_HOST
evals/runs/2026-09-26T11-43-49Z-none/outputs/command-gate/with-9/gate.py, line 116
with urllib.request.urlopen(request, timeout=REQUEST_TIMEOUT_S) as response:
evals/runs/2026-09-26T11-43-49Z-official/outputs/command-gate/with-9/gate.py, line 99
"network connection (curl/wget upload, scp/rsync to a "
scripts/install_muse.sh, line 34
curl -fsSL "https://codeload.github.com/$REPO/tar.gz/refs/heads/main" | tar -xz -C "$work"
Show more
evals/runs/2026-09-26T11-43-49Z-none/aggregate-result.json, line 780
"evidence": "\"\"\"Command approval gate backed by typesafe.ai's Jev model.\n\nGiven a shell command an agent wants to run and the agent's own stated\njustification for it, decides
evals/runs/2026-09-26T11-43-49Z-none/aggregate-result.json, line 818
"evidence": "\"\"\"Command approval gate backed by typesafe.ai's Jev model.\n\nGiven a shell command an agent wants to run and the agent's own stated\njustification for it, decides
evals/runs/2026-09-26T11-43-49Z-none/aggregate-result.json, line 832
"evidence": "\"\"\"Command approval gate backed by typesafe.ai's Jev model.\n\nGiven a shell command an agent wants to run and the agent's own stated\njustification for it, decides
Secrets
Concern
reads a credential or environment file.
SECRETS_ENV_READ SECRETS_PATH_CREDENTIAL SECRETS_PATH_DOTENV
evals/container/run.sh, line 17
creds=~/.claude/.credentials.json
evals/lib/provenance.sh, line 36
re.escape(os.environ["CLAUDE_CODE_OAUTH_TOKEN"]) if os.environ.get("CLAUDE_CODE_OAUTH_TOKEN") else "",
evals/runs/2026-09-26T11-43-49Z-none/outputs/log-labeler/with-1/label_logs.py, line 89
if not os.environ.get("TYPESAFE_API_KEY"):
Show more
evals/runs/2026-09-26T11-43-49Z-none/outputs/log-labeler/with-8/label_logs.py, line 132
if not os.environ.get("TYPESAFE_API_KEY"):
evals/runs/2026-09-26T11-43-49Z-official/outputs/command-gate/with-10/gate.py, line 38
"(e.g. SSH keys, tokens, .env files, sending them over the network), "
evals/runs/2026-09-26T11-43-49Z-official/outputs/command-gate/with-2/gate.py, line 67
"material (e.g. cat-ing `.env`, dumping keychains, printing cloud "
evals/runs/2026-09-26T11-43-49Z-official/aggregate-result.json, line 221
"evidence": "\"\"\"Auto-approval gate for coding-agent shell commands, backed by TypeSafe's Jev model.\n\nThe agent proposes a `command` and writes its own `agent_reason` for wanti
evals/runs/2026-09-26T11-43-49Z-official/aggregate-result.json, line 259
"evidence": "\"\"\"Auto-approval gate for coding-agent shell commands, backed by TypeSafe's Jev model.\n\nThe agent proposes a `command` and writes its own `agent_reason` for wanti
evals/runs/2026-09-26T11-43-49Z-official/aggregate-result.json, line 273
"evidence": "\"\"\"Auto-approval gate for coding-agent shell commands, backed by TypeSafe's Jev model.\n\nThe agent proposes a `command` and writes its own `agent_reason` for wanti
evals/runs/2026-09-26T11-43-49Z-official/aggregate-result.json, line 861
"evidence": "\"\"\"Command gate for a coding agent, backed by TypeSafe's Jev model.\n\n`review(command, agent_reason)` decides whether a shell command the agent\nwants to run shoul
evals/runs/2026-09-26T11-43-49Z-official/aggregate-result.json, line 899
"evidence": "\"\"\"Command gate for a coding agent, backed by TypeSafe's Jev model.\n\n`review(command, agent_reason)` decides whether a shell command the agent\nwants to run shoul
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
evals/runs/2026-09-26T11-43-49Z-none/aggregate-result.json, line 1
{
evals/runs/2026-09-26T11-43-49Z-official/aggregate-result.json, line 1
{
evals/runs/2026-09-26T11-43-49Z-this/aggregate-result.json, line 1
{
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 a46856c9fbc8. requests shell access. The manifest names this permission. install fetches a script from the network. README documents an installer that pipes a download into a shell. The docs document an installer that pipes a download into a shell. can reach a host that is not written as a fixed address. can reach codeload.github.com. The description does not name that host. reads a credential or environment file. a file was over the read cap and was not fully read.
Date and source
Date checked:
github.com
Commit graded: a46856c9fbc890653891e25f85ee2f16e1d9b513
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.