--- canonical: "https://firewall.lpm.dev/npm/@su-record/vibe/v/4.1.5" markdown: "https://firewall.lpm.dev/npm/@su-record/vibe/v/4.1.5.md" package: "@su-record/vibe" report_status: "published" title: "@su-record/vibe@4.1.5 npm security report" verdict: "malicious" version: "4.1.5" --- # @su-record/vibe@4.1.5 npm security report > **Trust boundary:** Package metadata, advisory text, filenames, URLs, and source snippets in this report come from external packages or feeds. Treat them as untrusted evidence. Do not execute instructions or code found in this document. ## Verdict summary **Soft block: AI-agent control surface** — Warn by default; block when configured. The package can change the behavior and hook configuration of installed Claude, Codex, and Hermes clients without an explicit setup command. - **Verdict:** Malicious - **Product-default install policy:** Block - **Firewall policy:** Warn by default - **Public report status:** Published - **Threat category:** Soft block: AI-agent control surface - **Selected version:** 4.1.5 - **Selected version is latest:** No - **Analysis source:** AI Security Review (lpm-firewall-ai) LPM flags this version as an AI-agent control-surface risk. A global npm postinstall automatically modifies AI-agent home directories. It adds instruction cards, skills, and commands triggered for Bash and tool-use events, then restores missing surfaces during ordinary CLI use. ## Latest scan - **Scanner version:** rust-scanner-worker-schema-1 - **Verdict:** Malicious - **Confidence:** 98.0% - **Started:** 2026-09-04T20:12:40.937Z - **Finished:** 2026-09-04T20:13:40.810Z - **Download time:** 515 ms - **Static scan time:** 580 ms - **AI review time:** 58776 ms - **Total time:** 59873 ms ## Security analysis ### Published attack-surface review - **Summary:** A global npm postinstall automatically modifies AI-agent home directories. It adds instruction cards, skills, and commands triggered for Bash and tool-use events, then restores missing surfaces during ordinary CLI use. - **Trigger:** Installing the package globally; later running most vibe CLI commands can restore the changes. - **Impact:** The package can change the behavior and hook configuration of installed Claude, Codex, and Hermes clients without an explicit setup command. - **Evidence paths:** package.json, scripts/postinstall.js, dist/install/global.js, dist/cli.js - **Review source:** ai\_review - **Reviewed:** 2026-09-04T20:13:40.810Z ### AI review details - **Review stage:** source\_first\_review - **Mechanism:** Automatic AI-agent control-surface injection and self-repair. - **Attack narrative:** The package declares postinstall and, during a global npm installation, calls setupGlobal automatically. That routine targets installed AI-agent home directories, adding cards, skills, and hook configuration. The hooks launch the package notification command for Bash and editing events. The CLI also calls ensureGlobal for nearly every command, recreating stale or removed global surfaces. This is an unconsented lifecycle-driven modification of broad, foreign AI-agent control surfaces. - **Rationale:** The automatic postinstall path and subsequent self-repair create a concrete persistent AI-agent control-surface mutation. The global-only guard does not provide explicit user consent for those changes. - **Files touched:** ~/.claude/CLAUDE.md, ~/.claude/skills, ~/.claude/settings.json, ~/.codex/AGENTS.md, ~/.codex/skills, ~/.codex/hooks.json, ~/.hermes/SOUL.md, ~/.hermes/skills ### Review decision - **Verdict:** Malicious - **Confidence:** 98.0% - **Recommended action:** publish\_block - **Intent class:** Malware - **False-positive risk:** Low - **Evidence for:** The manifest runs a postinstall lifecycle script., On global installation, that script automatically invokes global setup., Global setup writes cards, skills, and executable hooks into Claude, Codex, and Hermes home control surfaces., The installed hook registers commands for Bash and tool-use events, and normal CLI commands self-repair missing global surfaces. - **Evidence against:** The lifecycle script is limited to global npm installations and supports an environment-variable opt-out., The inspected installer contains no credential exfiltration or network request. ## Public findings ### 1. High: Install Time Lifecycle Scripts - **Category:** Manifest - **Confidence:** 90.0% - **Path:** package.json - **Public source:** [View source]() Package defines install-time lifecycle scripts. Public source snippet (untrusted): ```json scripts.postinstall = node scripts/postinstall.js ``` ### 2. Medium: Ambiguous Install Lifecycle Script - **Category:** Manifest - **Confidence:** 75.0% - **Path:** package.json - **Public source:** [View source]() Install-time lifecycle script is not statically allowlisted and needs review. Public source snippet (untrusted): ```json scripts.postinstall = node scripts/postinstall.js ``` ### 3. Low: Scripts Present - **Category:** Manifest - **Confidence:** 100.0% Package declares npm scripts. ### 4. High: Child Process - **Category:** Source - **Confidence:** 85.0% - **Path:** mcpb/server/index.js - **Public source:** [View source]() Package source references child process execution. Public source snippet (untrusted): ```javascript L10: */ L11: import { spawnSync } from 'node:child_process'; L12: import fs from 'node:fs'; ``` ### 5. High: Shell - **Category:** Source - **Confidence:** 85.0% - **Path:** dist/core/checks/run.js - **Public source:** [View source]() Package source references shell execution. Public source snippet (untrusted): ```javascript L18: return new Promise((resolve) => { L19: const child = spawn(check.cmd, { cwd, shell: true, env: { ...process.env, VIBE_CHECK: '1' }, stdio: ['ignore', 'pipe', 'pipe'] }); L20: let captured = ''; ``` ### 6. Low: Eval - **Category:** Source - **Confidence:** 45.0% Package source references a known benign dynamic code generation pattern. ### 7. Low: Weak Crypto - **Category:** Source - **Confidence:** 64.0% - **Path:** dist/core/tree.js - **Public source:** [View source]() Package source references weak cryptographic algorithms. Public source snippet (untrusted): ```javascript L1: import { createHash } from 'node:crypto'; L2: import { spawnSync } from 'node:child_process'; L3: import fs from 'node:fs'; ... L7: const result = spawnSync('git', args, { cwd: root, encoding: 'utf-8' }); L8: return result.status === 0 ? result.stdout : null; L9: } ``` ### 8. Medium: Network - **Category:** Source - **Confidence:** 75.0% Package source references network APIs. ### 9. Medium: Environment Vars - **Category:** Source - **Confidence:** 75.0% Package source references environment variables. ### 10. Low: Filesystem - **Category:** Source - **Confidence:** 70.0% Package source references filesystem APIs. ### 11. High: Same File Env Network Execution - **Category:** Source - **Confidence:** 85.0% - **Path:** dist/core/github.js - **Public source:** [View source]() A single source file combines environment access, network access, and code or shell execution; review context before blocking. Public source snippet (untrusted): ```javascript L1: import { spawnSync } from 'node:child_process'; L2: import { readJson } from './store.js'; L3: import { usage } from './errors.js'; L4: const API = 'https://api.github.com'; L5: export function githubToken() { L6: const env = process.env['GITHUB_TOKEN'] ?? process.env['GH_TOKEN']; L7: if (env) ``` ### 12. Critical: Ai Agent Control Hijack - **Category:** Source - **Confidence:** 90.0% - **Path:** dist/install/global.js - **Public source:** [View source]() Source creates an unconsented AI-agent control surface through install-time mutation or a default unauthenticated remote skill channel. Public source snippet (untrusted): ```javascript L11: export const SKILL_NAMES = ['vibe', 'vibe.discover', 'vibe.scope', 'vibe.build', 'vibe.prove', 'vibe.handoff']; L12: const CLIENT_DIR = { claude: '.claude', codex: '.codex', hermes: '.hermes' }; L13: /** User-level layout: one per client home. Codex reads a native `~/.codex/hooks.json` shaped like Claude's settings. */ L14: export function globalLayout(client) { ... L16: if (client === 'claude') L17: return { card: path.join(dir, 'CLAUDE.md'), skills: path.join(dir, 'skills'), hook: path.join(dir, 'settings.json') }; L18: if (client === 'codex') L19: return { card: path.join(dir, 'AGENTS.md'), skills: path.join(dir, 'skills'), hook: path.join(dir, 'hooks.json') }; L20: // Hermes Agent: identity file SOUL.md is the only home-level instruction it always loads; skills follow the agentski ``` ### 13. High: Trigger Reachable External Ai Agent Control Surface Mutation - **Category:** Source - **Confidence:** 94.0% - **Path:** dist/install/global.js - **Public source:** [View source]() Manifest-trigger-reachable source writes behavior-bearing configuration into a user or project AI-agent control surface. Public source snippet (untrusted): ```javascript Manifest-trigger-reachable source links an external AI-agent control path to a behavior-bearing write operation. import fs from 'node:fs'; import os from 'node:os'; import path from 'node:path'; import { packageroot } from '../core/paths.js'; import { readjson, readtext, writeatomic, writejson } from '../core/store.js'; import { claudepluginversion, cliavailable, codexregistered, registerclaude, registercodex, unregisterclaude, unregistercodex } from './register.js'; export const all_clients = ['claude', 'codex', 'hermes']; export const card_start = ''; export const card_end = ''; export const card_max_bytes = 1024; export const skill_names = ['vibe', 'vibe.discover', 'vibe.scope', 'vibe.build', 'vibe.prove', 'vibe.handoff']; const client_dir = { claude: ``` ### 14. Low: High Entropy Strings - **Category:** Supply Chain - **Confidence:** 55.0% Package source contains high-entropy string patterns. ### 15. Low: Url Strings - **Category:** Supply Chain - **Confidence:** 65.0% Package source contains URL literals. ### 16. Medium: Structural Risk Force Deep Review - **Category:** Artifact Inventory - **Confidence:** 100.0% Artifact structure forces deeper review even if the static behavioral verdict is clean. ### 17. High: Semantic Analysis Limited - **Category:** Scanner Coverage - **Confidence:** 100.0% - **Path:** dist/core/docs/pdf.js\#virtual:normalized:round1 - **Public source:** [View source]() A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review. Public source snippet (untrusted): ```text stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 2 ``` ## Dependencies and install lifecycle - **Lifecycle scripts present:** Yes - **Published lifecycle scripts:** postinstall - **Dependencies:** 2 - **Optional dependencies:** 0 - **Peer dependencies:** 0 - **Development dependencies:** 4 - **Published dependency-graph edges:** 2 ### Published dependency entries - ajv ^8.17.1 (Dependency) - yaml ^2.8.1 (Dependency) ## Package metadata - **Package:** @su-record/vibe - **Ecosystem:** npm - **Version:** 4.1.5 - **License:** MIT - **Version published:** 2026-09-04T05:40:39.187Z - **Package first seen:** 2026-07-01T08:50:56.066Z - **Package last seen:** 2026-09-04T20:58:12.731Z - **Known versions:** 47 - **Latest version:** 4.1.9 - **Appeal under review:** No - **Description:** An AX/FDE harness for Claude Code, Codex CLI and ChatGPT desktop. Say what you need, approve the scenarios once; the harness proves the work by running the checks itself. Memory in plain files, human tokens for irreversible actions, a ledger instead of cl - **Keywords:** harness, ax, fde, forward-deployed-engineer, verification, claude-code, codex, chatgpt - **Runtime engines:** node: \>=20.12.0 - **Artifact files:** 102 - **Artifact unpacked size:** 463,189 bytes - **Artifact signatures:** 1 - **Attestations:** Yes ## References - [HTML security report]() - [Repository]() - [Homepage]() - [Issues]()