Claude Code, Skill
skrun
skrun, a skill listed in the PlugVet directory.
Compatibility: Not stated.
Grade D
Two or more Medium findings, or a High finding. Only try it in a sandbox.
Not a guarantee. A hygiene check of the source at the commit shown, not a security audit.
Rescanned at commit b1d963b36174, unchanged since the check below.
PERMISSION_SHELL_UNDECLARED INSTALL_REMOTE_SCRIPT
Why this grade
- PERMISSION_SHELL_UNDECLARED Medium
Runs shell commands that the manifest or description does not mention.
packages/runtime/src/tools/script-deps-installers.ts, line 73
/** Production runner - wraps `node:child_process.execFile`. */ - 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).
infra/runtime-image/Dockerfile, line 91
# Was `curl https://mise.run | sh` - an unpinned remote
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
Concern
requests shell access. No manifest field names shell or command execution.
PERMISSION_SHELL_UNDECLARED
packages/runtime/src/tools/script-deps-installers.ts, line 73
/** Production runner - wraps `node:child_process.execFile`. */
packages/runtime/src/tools/script-deps-resolver.ts, line 26
* Defaults to `execFileRunner` (real `child_process.execFile`).
Install and update
Serious
install fetches a script from the network.
INSTALL_REMOTE_SCRIPT
CHANGELOG.md, line 121
remote installer script into a shell (`curl https://mise.run | sh`) with no version, checksum or signatur
infra/runtime-image/Dockerfile, line 91
# Was `curl https://mise.run | sh` - an unpinned remote
Network
Concern
can reach a host that is not written as a fixed address. can reach example.com. The description does not name that host. can reach evil.invalid. The description does not name that host. can reach 1.1.1.1. The description does not name that host. can reach cdn.jsdelivr.net. The description does not name that host.
NETWORK_NO_FIXED_HOST NETWORK_UNNAMED_HOST
agents/adr-writer/README.md, line 30
curl -X POST http://localhost:4000/api/agents/dev/adr-writer/run \
agents/adr-writer/README.md, line 48
curl http://localhost:4000/api/runs/<run_id>/files/0003-*.md \
agents/changelog-generator/README.md, line 32
# 3. Call it via the CLI (or use curl below)
Show more
agents/changelog-generator/README.md, line 37
curl -X POST http://localhost:4000/api/agents/dev/changelog-generator/run \
agents/changelog-generator/README.md, line 51
curl http://localhost:4000/api/runs/<run_id>/files/CHANGELOG.md \
agents/changelog-generator/README.md, line 53
curl http://localhost:4000/api/runs/<run_id>/files/release-notes.md \
agents/csv-to-executive-report/README.md, line 37
curl -X POST http://localhost:4000/api/agents/dev/csv-to-executive-report/run \
agents/csv-to-executive-report/README.md, line 52
curl http://localhost:4000/api/runs/<run_id>/files/report.pdf \
agents/knowledge-base-from-vault/README.md, line 34
curl -X POST http://localhost:4000/api/agents/dev/knowledge-base-from-vault/run \
agents/semgrep-rule-creator/README.md, line 36
"cve_description": "Server-Side Request Forgery via user-controlled URL passed directly to fetch() without allowlist validation.",
agents/semgrep-rule-creator/README.md, line 37
"bad_code_example": "const url = req.query.url; const res = await fetch(url); res.send(await res.text());",
agents/semgrep-rule-creator/fixtures/sample-cve.md, line 31
const result = await fetch(url);
agents/semgrep-rule-creator/fixtures/sample-cve.md, line 48
const result = await fetch(url);
infra/runtime-image/entrypoint.sh, line 94
# Without this rule, getent / curl can't resolve the allowed hosts.
infra/runtime-image/zero-trust.test.sh, line 78
# we curl from a debug entrypoint rather than the real runner-start.
infra/runtime-image/zero-trust.test.sh, line 89
# debug probe (skips the capsh handoff so we can curl as root after).
Secrets
Concern
reads a credential or environment file.
SECRETS_ENV_READ SECRETS_PATH_DOTENV
infra/runtime-image/runner/src/index.ts, line 130
if (isPoolMode() && !process.env.RUNNER_CLAIM_TOKEN) {
infra/runtime-image/runner/src/index.ts, line 157
claimToken: process.env.RUNNER_CLAIM_TOKEN,
infra/runtime-image/runner/src/index.ts, line 158
runToken: process.env.RUNNER_RPC_TOKEN,
Show more
infra/runtime-image/runner/src/index.ts, line 249
process.env.RUNNER_RPC_TOKEN = parsed.data.rpcToken;
infra/runtime-image/runner/src/index.ts, line 255
delete process.env.RUNNER_CLAIM_TOKEN;
packages/api/src/index.test.ts, line 140
delete process.env.FLY_API_TOKEN;
packages/api/src/index.test.ts, line 142
delete process.env.S3_ACCESS_KEY_ID;
packages/api/src/index.test.ts, line 143
delete process.env.S3_SECRET_ACCESS_KEY;
packages/api/src/index.ts, line 269
"See .env.example.",
packages/api/src/runtime/adapter-selection.ts, line 70
"See .env.example for the supported values.",
packages/api/src/runtime/adapter-selection.ts, line 107
"See .env.example for the full cloud-runtime configuration.",
packages/api/src/utils/webhook.ts, line 32
"Set the WEBHOOK_SIGNING_KEY environment variable (see .env.example).",
packages/cli/src/commands/build.test.ts, line 39
expect(AGENT_BUNDLE_EXCLUDES.has(".env")).toBe(true);
packages/runtime/src/tools/script-deps-installers.test.ts, line 363
expect(uvCall?.env?.UV_PROJECT_ENVIRONMENT).toBe(join(depsPath, "venv"));
packages/runtime/src/tools/script-deps-installers.test.ts, line 364
expect(uvCall?.env?.UV_INDEX_URL).toBe(PYPI_INDEX_URL);
packages/runtime/src/tools/script-deps-installers.test.ts, line 450
expect(poetryCall?.env?.POETRY_VIRTUALENVS_CREATE).toBe("false");
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
agents/meeting-transcript-to-action-items/fixtures/sample.wav, line 1
RIFF�} WAVEfmt �>