Loading npm security reports…

sprag-cli@3.54.0: AI-agent control-surface risk | LPM Firewall

registry / sprag-cli / 3.54.0

sprag-cli@3.54.0

Sprag - a local harness for Claude Code and Codex: ratchet rules, model delegation, token telemetry, companion panels, document conversion, and Korean writing checks.

AI Security Review

scanned 10d ago · by lpm-firewall-ai

LPM flags this version as an AI-agent control-surface risk. An npm postinstall automatically configures a broad Claude Code control surface. No user command or interactive approval is required in unattended installation.

Static reason

High-risk behavior combination matched malicious policy.; source matched previously finalized malicious package; routed for review

Trigger

Installing the package from npm triggers postinstall.

Impact

The package can change future Claude Code behavior across projects and cause package commands to run during agent events.

Mechanism

The lifecycle hook runs the install command, which writes global agent integration assets and registers session and tool hooks.

Policy narrative

The package declares a postinstall hook that runs its installer automatically. That installer configures a global Claude Code skill, statusline, session hooks, tool hooks, a global harness block, and document-conversion hooks. The install flow explicitly falls back to automatic defaults when no interactive terminal is available, which is the normal lifecycle-install context. This is an unconsented install-time mutation of a broad AI-agent control surface.

AI rationale

The automatic lifecycle path modifies global Claude Code configuration and registers future agent hooks without consent. This meets the install-control-surface blocking boundary.

Evidence

package.jsonsrc/commands/install.jssrc/installer.js~/.claude/CLAUDE.md~/.claude/settings.json~/.claude/skills/claude-token-saver/SKILL.md

Decision evidence

Source & flagged code

19 flagged · loading source

package.jsonView file

•scripts.postinstall = node bin/cli.js install || true

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg

•scripts.postinstall = node bin/cli.js install || true

Medium

Ambiguous Install Lifecycle Script

Install-time lifecycle script is not statically allowlisted and needs review.

package.jsonView on unpkg

src/doc2md.cjsView file

28const crypto = require('node:crypto'); L29: const { spawn, spawnSync } = require('node:child_process'); L30: const ledger = require('./doc2md-ledger.cjs');

High

Child Process

Package source references child process execution.

src/doc2md.cjsView on unpkg · L28

24L25: const fs = require('node:fs'); L26: const os = require('node:os');

Medium

Dynamic Require

Package source references dynamic require/import behavior.

src/doc2md.cjsView on unpkg · L24

src/fig2md.cjsView file

31// `0.4.x` rather than `^0.4.1` on purpose: the Windows install goes through L32: // cmd.exe, where `^` is the escape character and would be eaten before npm L33: // ever saw it. The range is the same one caret means for a 0.x package.

High

Shell

Package source references shell execution.

src/fig2md.cjsView on unpkg · L31

src/seed-rules.jsView file

56try { L57: const data = JSON.parse(readFileSync(seedStatePath(opts), 'utf8')); L58: return data && typeof data.decided === 'object' && data.decided ? data : { decided: {} };

Low

Weak Crypto

Package source references weak cryptographic algorithms.

src/seed-rules.jsView on unpkg · L56

src/codex-panel-shell.jsView file

34Trigger-reachable persistence chain: scripts.postinstall -> bin/cli.js -> src/commands/codex.js -> src/codex-panel-shell.js L34: L35: export function installPanelShell({ file = join(process.env.ZDOTDIR || homedir(), '.zshrc'), remove = false } = {}) { L36: const text = existsSync(file) ? readFileSync(file, 'utf8') : '';

High

Trigger Reachable Persistence

A manifest entrypoint or package-local install chain reaches persistence behavior.

src/codex-panel-shell.jsView on unpkg · L34

34L35: export function installPanelShell({ file = join(process.env.ZDOTDIR || homedir(), '.zshrc'), remove = false } = {}) { L36: const text = existsSync(file) ? readFileSync(file, 'utf8') : '';

Medium

Install Persistence

Source writes installer persistence such as shell profile or service configuration.

src/codex-panel-shell.jsView on unpkg · L34

src/commands/feedback.jsView file

56if (auth.status !== 0) return null; L57: const res = spawnSync('gh', ['issue', 'create', '-R', REPO, '--title', title, '--body', body], L58: { encoding: 'utf8', timeout: 30_000 }); L59: if (res.status !== 0) { debug('feedback:gh', res.stderr); return null; } L60: const url = String(res.stdout).trim().split('\n').pop(); ... L67: try { L68: const params = new URLSearchParams(); L69: params.set(FORM.message, `${title}\n\n${body}`); L70: params.set(FORM.meta, meta); L71: const res = await fetch(`https://docs.google.com/forms/d/e/${FORM.id}/formResponse`, { L72: method: 'POST',

High

Command Output Exfiltration

Source combines command execution, command-output handling, and outbound requests; review data flow before blocking.

src/commands/feedback.jsView on unpkg · L56

21L22: import { spawnSync } from 'node:child_process'; L23: import { writeFileSync, mkdirSync } from 'node:fs'; ... L28: const REPO = 'rootstudioyaml/sprag'; L29: const ISSUES_URL = `https://github.com/${REPO}/issues`; L30: ... L47: `version: ${version}`, L48: `os: ${process.platform} ${os.release()}`, L49: `node: ${process.version}`, ... L58: { encoding: 'utf8', timeout: 30_000 }); L59: if (res.status !== 0) { debug('feedback:gh', res.stderr); return null; } L60: const url = String(res.stdout).trim().split('\n').pop();

High

Sandbox Evasion Gated Capability

Source gates dangerous network, credential, or execution behavior behind CI, host, platform, time, or geo fingerprint checks.

src/commands/feedback.jsView on unpkg · L21

56Trigger-reachable command-output exfiltration chain: scripts.postinstall -> bin/cli.js -> src/commands/feedback.js L56: if (auth.status !== 0) return null; L57: const res = spawnSync('gh', ['issue', 'create', '-R', REPO, '--title', title, '--body', body], L58: { encoding: 'utf8', timeout: 30_000 }); L59: if (res.status !== 0) { debug('feedback:gh', res.stderr); return null; } L60: const url = String(res.stdout).trim().split('\n').pop(); ... L67: try { L68: const params = new URLSearchParams(); L69: params.set(FORM.message, `${title}\n\n${body}`); L70: params.set(FORM.meta, meta); L71: const res = await fetch(`https://docs.google.com/forms/d/e/${FORM.id}/formResponse`, { L72: method: 'POST',

bin/cli.jsView file

6Install-time code directly mutates a foreign AI-agent control surface: L119: --project <path> / --session <id> / --interval 2 / --once / --no-color L120: sprag install --agent codex install Codex hooks + AGENTS.md harness L121: sprag harness init --agent codex initialize this project's Codex harness ... L128: sprag uninstall remove everything install added L129: sprag harness check score the harness setup in CLAUDE.md L130: sprag harness analyze run the harness transcript analysis manually ... L287: L288: // Subcommand: harness — manage the project's CLAUDE.md harness rules. L289: // sprag harness init # write CLAUDE.md (5 sections) + ratchet.md L290: // sprag harness uninit # remove harness block from CLAUDE.md (backup kept) L291: // sprag harness che

Critical

Ai Agent Control Hijack

Source creates an unconsented AI-agent control surface through install-time mutation or a default unauthenticated remote skill channel.

bin/cli.jsView on unpkg · L6

examples/statusline-command.ps1View file

•path = examples/statusline-command.ps1 kind = build_helper sizeBytes = 1918 magicHex = [redacted]

Medium

Ships Build Helper

Package ships non-JavaScript build or shell helper files.

examples/statusline-command.ps1View on unpkg

src/advice.js#virtual:normalized:round1View file

•stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 5

High

Semantic Analysis Limited

A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.

src/advice.js#virtual:normalized:round1View on unpkg

src/korean-style.jsView file

•matchType = normalized_sha256 matchedPackage = claude-token-saver@3.50.0 matchedPath = src/korean-style.js matchedIdentity = npm:Y2xhdWRlLXRva2VuLXNhdmVy:3.50.0 similarity = 1.000 summary = normalized source hash matched finalized malicious source

High

Known Malware Source Similarity

Source file is highly similar to a previously finalized malicious package; route for source-aware review.

src/korean-style.jsView on unpkg

src/gateway-auth.jsView file

•matchType = normalized_sha256 matchedPackage = claude-token-saver@3.50.0 matchedPath = src/gateway-auth.js matchedIdentity = npm:Y2xhdWRlLXRva2VuLXNhdmVy:3.50.0 similarity = 1.000 summary = normalized source hash matched finalized malicious source

High

Known Malware Source Similarity

Source file is highly similar to a previously finalized malicious package; route for source-aware review.

src/gateway-auth.jsView on unpkg

src/litellm-budget.jsView file

•matchType = normalized_sha256 matchedPackage = claude-token-saver@3.50.0 matchedPath = src/litellm-budget.js matchedIdentity = npm:Y2xhdWRlLXRva2VuLXNhdmVy:3.50.0 similarity = 1.000 summary = normalized source hash matched finalized malicious source

High

Known Malware Source Similarity

Source file is highly similar to a previously finalized malicious package; route for source-aware review.

src/litellm-budget.jsView on unpkg

src/model-alias.jsView file

•matchType = normalized_sha256 matchedPackage = claude-token-saver@3.50.0 matchedPath = src/model-alias.js matchedIdentity = npm:Y2xhdWRlLXRva2VuLXNhdmVy:3.50.0 similarity = 1.000 summary = normalized source hash matched finalized malicious source

High

Known Malware Source Similarity

Source file is highly similar to a previously finalized malicious package; route for source-aware review.

src/model-alias.jsView on unpkg