More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Safe npm install: evidence before dependencies

    0

    A dependency decision playbook for agents: check releases, explain risk, pin versions, and review install scripts.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Safe npm install: evidence before dependencies

    Safe npm install: evidence before dependencies

    Example session with this skill installed

    Add axios@1.14.1 to this project. Check the release with Dep Bouncer first, explain the evidence, and if it is blocked, suggest the safe version and the exact install command without installing anything yet.

    • Read your context and instructions
    • Compiled the safe npm install

    Summary

    BLOCK: Do not install axios@1.14.1. The replay reports a newly added suspicious dependency, a provenance drop, a publisher change, and a very recent publication.

    Finding titleSignalWeight / detail
    New dependency plain-crypto-js looks like a malware dropnew-depsWeight 70; added as ^4.2.1, runs postinstall: node setup.js, has no importing file, and resembles crypto-js.
    Provenance dropped: previous release 1.14.0 was built and signed by CI, this one was notprovenance-dropWeight 35; axios@1.14.1 has no provenance attestation; 4 of the previous 5 releases had one.
    Published by the personal account "jasonsaayman" instead of the package's CI (trusted publishing)publisher-changeWeight 35; the last 4 releases were published by GitHub Actions, while 1.14.1 was published by jasonsaayman.
    Published 30 min agorelease-timingWeight 15; the output says a 24 h cooldown would hold this release.

    The server's suggested historical safe version is

    1.14.0 (the last release before the hijack). Check that exact version live before installation; a replay suggestion is historical, not current installation authorization. If that check permits the install, the proposed exact command is:

    npm install axios@1.14.0 --save-exact

    Next steps

    • Do not install axios@1.14.1.
    • If you want the alternative, check axios@1.14.0 live and review its current result before installing.
    • Keep npm audit and project validation; the replay notes that npm audit found 0 vulnerabilities because no advisory was yet available at that time.

    Evidence is a replay of the March 2026 axios hijack using the server's labelled incident fixture. The malicious version was removed from npm; the bad release is reconstructed with benign stand-in code.

    
    Optional ALLOW example from the captured live output:
    
    

    chalk@6.0.1 returned ALLOW with score 0 and no findings: “looks consistent with its own history. No risk signals.” Proceed only when installation is authorized, and pin the checked version exactly.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Coding agents can change dependencies faster than a person can inspect a release. Safe npm install puts a consistent decision process before npm, pnpm, yarn, and bun changes and package-safety questions. It uses the free Dep Bouncer MCP server to inspect release history, publisher and provenance changes, install behavior, dependencies, code findings, and advisory evidence.

    The agent playbook complements the free server and its bundled prompt: it reconciles incomplete coverage, separates range audits from locked versions, checks peers, explains evidence plainly, and pins the exact checked version. ALLOW permits an authorized install; REVIEW asks the user; BLOCK prevents that release and offers a checked alternative when available. Script decisions remain specific to the package and version.

    For Claude Code, Cursor, and generic Streamable-HTTP MCP agents, the package includes setup instructions, a finding reference, and worked examples. It is readable Markdown with no hidden execution, telemetry, credential handling, or bundled enforcement hook.

    What you get

    • Separate workflows for adds, upgrades, audits, scripts, version selection, and safety questions.
    • BLOCK stops and incomplete or uncertain checks require review.
    • Minimal manifests are built locally; skipped entries and unaudited peers are reconciled.
    • Checked versions are pinned exactly and unnecessary lifecycle scripts stay disabled.
    • Signal and code-pattern evidence is translated into practical explanations.

    How it works

    1. Connect a skill-capable agent to the free hosted Dep Bouncer MCP server or a configured self-hosted server.
    2. Choose the workflow for the requested add, upgrade, audit, script review, version selection, or safety question.
    3. Send only the tool arguments required for that check; for a manifest audit, first build the allowed minimal manifest locally.
    4. Read the structured result, findings, completeness, and coverage notes, then apply ALLOW, REVIEW, or BLOCK rules.
    5. For an authorized install, pin the exact checked version, preserve the project’s package manager, and report actual changes and warnings.

    Privacy and network calls

    MCP calls go to the configured endpoint, by default the author's free hosted service. The server receives package names and requested versions/ranges, or the specific upgrade, script-review, incident, or minimal-manifest arguments for the selected tool, plus ordinary MCP/HTTP metadata. A minimal manifest may disclose the project's name/version and dependency names, values, and section labels; scripts and other project fields are excluded. Private or non-registry dependencies are omitted and treated as unresolved REVIEW. Live checks can query npm registry metadata/tarballs, npm download counts, and OSV advisories. Replays use bundled reconstruction fixtures without live registry/OSV lookups. Installs make normal package-manager requests; approved scripts can make additional calls. The playbook adds no telemetry or uploads of source files, lockfiles, conversation text, environment, or credentials. Checks are npm heuristics, not a safety guarantee.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 28 days ago

    Requires

    Nitrocloud
    Dep Bouncer MCP
    Node.js
    npm or pnpm or Yarn or Bun

    Permissions

    Terminal / Shell
    Read Files
    Network Access
    Write Files

    Allowed Hosts

    nitrocloud.ai

    File Scopes

    package.json
    package-lock.json
    pnpm-lock.yaml
    yarn.lock
    bun.lock*

    Read package.json and the applicable lockfile locally to identify dependencies and locked versions; the playbook builds a minimal manifest before a manifest check. Network access is needed for MCP discovery and checks against the configured Dep Bouncer endpoint; live checks may cause the server to query npm registry metadata/tarballs, npm download counts, and OSV advisories. The server receives the selected tool's package names and version/range arguments; manifest checks may also disclose project name/version, dependency keys and values, and section labels. Scripts, other project fields, source files, lockfiles, conversation text, environment, and credentials are not uploaded by this playbook. Private/non-registry dependency specs are omitted locally and reported as unresolved REVIEW. Actual authorized installs use the project's normal package-manager registry/artifact and configured audit requests; approved lifecycle scripts may make their own network calls.

    Listed28 days ago

    What's inside

    Frequently Asked Questions

    Yes. The hosted server is free and run by the skill author; self-hosting is supported.

    The chosen MCP tool receives its listed arguments, such as package name and version/range. A manifest check sends a locally built minimal JSON manifest and includeDev; that can disclose project name/version and dependency names, values, and section labels. Ordinary MCP/HTTP metadata also accompanies requests. The playbook does not send source files, lockfiles, conversation text, environment, or credentials.

    Unavailable, timed-out, malformed, or incomplete checks are REVIEW, never ALLOW. Pause dependency changes until the check is complete or the connection is restored.

    No. It adds heuristic release and code evidence; keep normal vulnerability checks and project validation. Its checks are not a guarantee and do not cover every vulnerability, locked/transitive dependency, or malicious release.

    The workflows cover npm, pnpm, yarn, and bun. Use the existing manager and its appropriate exact-save and script-suppression options.

    New in Security