Skills/npm/jevelry

jevelry@0.7.0

VERIFIED· install only

Publisher claims: "Use Jev everywhere to make & track decisions" Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry npm (MCP server)Publisher backant-io · publishing for 53 daysInstalls 71Last detonated 45 h ago
$claude mcp add jevelry -- npx -y jevelry
WatchReport an issue

Coverage

Nothing executable in the bundle ran; the trace covers the install step only.
Tested
every file read, file written and host contacted by what ran
Not tested
use of the skill by an agent

Is [email protected] safe to install?

KillerSkills ran it in the KillerSkills honeypot sandbox (1 scenario; nothing executable in the bundle ran; the trace covers the install step only) on 2026-11-12 and scored it 81/100 (Verified). In what ran, nothing we could detect went wrong: no undeclared network egress, no credential reads, no persistent instruction writes. What it does with a real credential was not observed; the static findings and the publisher carry the rest of the score. Summary: no network, touches nothing outside the project, writes no memory.

Behavior trace

Honeypot run · 0 s · 1 scenarios · 0 deviations
t+CategoryObserved action
0.0sprocessno install scripts in bundle
0.0sprocessscript backend runs setup only
0.0sprocessscript backend runs setup only
0.0sprocessno install scripts to re-run

Static findings

MEDIUM
File type combined with upload actionskills/jevelry/SKILL.md:43 · File type combined with upload action — agent may exfiltrate local data files… smaller or local one might do | `npx jevelry ask request-route --state @request.json` | `complexity` at level 0 and `needs_tools` no: send it to the small model | send it to the large model | send it to the large model | | you finished wo…Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
File type combined with upload actionskills/jevelry/SKILL.md:35 · File type combined with upload action — agent may exfiltrate local data files… a report to the CI owner or a fix | `npx jevelry run failing-test --state @test.json` | the shipped commands only print the next step for the `cause`: rerun with `retries`, report it to the CI owner, or fix the code; replace them with you…Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
HIGH
Access to sensitive file pathsdist/cli.js:3174 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filesreturn discoveryDirs({ cli: cli ?? [], ...env ? { env } : {}, cwd: process.cwd(), home: home() });Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
File type combined with upload actionskills/jevelry/SKILL.md:45 · File type combined with upload action — agent may exfiltrate local data files…s into the app you are building | `npx jevelry ask ticket-triage --state @ticket.json` | route it to the `team`, and `urgent` yes puts it on top of the queue | route it and flag it for the queue owner | leave it i…Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
npx invocation of non-well-known packageskills/jevelry/SKILL.md:14 · npx invocation of non-well-known package — may execute untrusted codenpx jevelry installFix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
File type combined with upload actionskills/jevelry/SKILL.md:47 · File type combined with upload action — agent may exfiltrate local data files… you are unsure whose it is | `npx jevelry ask escalation-route --state @request.json` | send it to the `route` it names | send it and copy the shared inbox | leave it for a person to route | | an alert fire…Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
npx invocation of non-well-known packageAGENTS.md:11 · npx invocation of non-well-known package — may execute untrusted codenpx jevelry models --localFix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
npx invocation of non-well-known packageskills/jevelry/SKILL.md:18 · npx invocation of non-well-known package — may execute untrusted code…oes that part alone. To run Jev on the machine instead of the hosted API, run `npx jevelry models --local`, which ranks the decision models on Hugging Face and says for e…Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
File type combined with upload actionskills/jevelry/SKILL.md:37 · File type combined with upload action — agent may exfiltrate local data files…nds on your pull request | `npx jevelry ask review-comment-kind --state @comment.json` | `defect`: fix the code, `environment`: fix or report the CI or the sandbox and leave the code alone, `style`: fold it into a cleanup pass | do the sam…Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
npx invocation of non-well-known packageAGENTS.md:14 · npx invocation of non-well-known package — may execute untrusted code…om then on every ask goes to a local `llama-server` that the first ask starts. `npx jevelry stop` stops it, and `npx jevelry install --typesafe` goes back to hosted Jev. T…Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
HIGH
Access to sensitive file pathsdist/chunk-K47OKAFG.js:9127 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filesconst { jevel: loaded } = loadJevel(name, discoveryDirs({ cli: options.jevels ?? [], ...env ? { env } : {}, cwd: process.cwd(), home }));Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
HIGH
Access to sensitive file pathsdist/chunk-K47OKAFG.js:8395 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filesconst { TYPESAFE_API_KEY: _key, ...env } = process.env;Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
npx invocation of non-well-known packageAGENTS.md:12 · npx invocation of non-well-known package — may execute untrusted codenpx jevelry install --local vllm-d3-flashFix: Pin versions and sources; vendor anything fetched at install time.
AST06
HIGH
Access to sensitive file pathsdist/chunk-K47OKAFG.js:8405 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filesenv: { ...env, JEVELRY_STATE: file, JEVELRY_DECISION: meta.decision, JEVELRY_OPTION: meta.option ?? "", JEVELRY_LOG_ID: meta.logId ?? "" }Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
npx invocation of non-well-known packageskills/jevelry/SKILL.md:16 · npx invocation of non-well-known package — may execute untrusted code…a terminal it changes nothing and names both ways. If you only need the skill, `npx jevelry install-skill` does that part alone. To run Jev on the machine instead of the …Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
npx invocation of non-well-known packageAGENTS.md:7 · npx invocation of non-well-known package — may execute untrusted codenpx jevelry installFix: Pin versions and sources; vendor anything fetched at install time.
AST06
LOW
Prompt injectionjevels/meeting-notes/JEVEL.md:84 · Prompt injection — known jailbreak technique reference · informational: not an instruction file- "Dan called the plan unworkable"Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01

Cost to the agent

Instructions loaded
≈5,457 tok
2 files · 80 in bundle

Tokens estimated at four characters each. Skills load their instruction text when triggered; the figure is what that costs.

Known vulnerabilities

No advisories in OSV for jevelry@0.7.0 as of 44 h ago. This is the package itself, not its dependencies.

jevelry@0.7.0 safety verdict: Verified (install only) — KillerSkills