Is Packmind MCP server safe?
- Possible command injection (exec with dynamic command)
- Agent or editor configuration runs a command or sends data automatically
- Python shell/command execution
What to do: Read the findings below before installing: each one opens the exact line of code it was found on.
Packmind MCP server is an MCP server analyzed by SkillTotal's deterministic static scanner. The scan found no malicious indicators, though 2 risky constructs are reported for review. It can: dynamic code execution, filesystem read, filesystem write, install time execution, mcp tools detected, network egress and shell execution — capabilities are what the code can do, not a verdict on intent. Risk score 30/100 (medium).
packmind 1.21.0
Get notified when this component changes
Change alerts aren't live yet. Leave your email and we'll let you know when they launch. We use it only for this and delete it on request.
Automated static-analysis result. It can contain false positives and false negatives, and is not a claim about the intent of Packmind MCP server's authors. Report a false positive.
Behavioral traits
How this component maps to the CSA agentic threat model. Descriptive — it never affects the risk score.
Findings (13)
The code builds an OS command out of values that can change at runtime, then runs it through a shell.
const stdout = execSync(`git ${cmd}`, {execSync(`npm install -g ${NPM_PACKAGE}@${version}`, {Why it matters: If any of those values come from untrusted input, an attacker can run their own commands on the machine.
Fix: Use execFile/spawn with an argument array instead of exec; never build a shell command string from external input.
The code turns strings into live code at runtime (eval / new Function / exec).
const func = new Function(
Why it matters: If those strings aren't fixed and trusted, they become a way to run arbitrary code.
Fix: Avoid evaluating dynamically constructed code; if unavoidable, ensure the input is a trusted constant and never derived from external data.
An MCP server entry launches a command on your host.
"command": "npx",
"command": "npx",
Why it matters: Trusting the manifest means running that binary — verify what it is and where it comes from.
Fix: Verify the launched command and its source before trusting this MCP server configuration.
The component can run operating-system commands or spawn processes.
const embedResult = await Bun.spawn(['bun', 'run', embedWasmScript], {const compileResult = await Bun.spawn(
import { execSync, ExecSyncOptions } from 'child_process';const stdout = execSync(`git ${cmd}`, {const result = execSync('git rev-parse --show-toplevel', {import { execSync } from 'child_process';execSync(`npm install -g ${NPM_PACKAGE}@${version}`, {import { spawnSync } from 'child_process';const result = spawnSync(editor, [tmpFile], {import { spawnSync } from 'node:child_process';const res = spawnSync(command, {import { execFileSync, spawnSync } from 'node:child_process';const claudeCheck = spawnSync('claude', ['--version'], { stdio: 'ignore' });import { spawnSync } from 'node:child_process';res = spawnSync('script', ['-q', tsFile, '/bin/sh', '-c', joined], {res = spawnSync('script', ['-qec', joined, tsFile], {const rsvg = spawnSync('rsvg-convert', ['-z', '2', svgPath, '-o', pngPath], {const ql = spawnSync(
import { execSync } from 'node:child_process';execSync('pnpm husky', { stdio: 'inherit' });execSync('pnpm chakra:typegen', { stdio: 'inherit' });Why it matters: Powerful and often legitimate — confirm the commands aren't built from untrusted input.
Fix: Confirm the command and its arguments are fully controlled and not derived from untrusted input; prefer execFile with an argument array.
The component can run operating-system commands or spawn processes.
result = subprocess.run(
["git", "-C", cwd, "symbolic-ref", "--short", "HEAD"],
capture_output=True, text=True, timeout=2
)Why it matters: Powerful and often legitimate — confirm the commands aren't built from untrusted input.
Fix: Confirm the command and its arguments are fully controlled and not derived from untrusted input; avoid shell=True.
The component ships agent or IDE configuration that acts without a separate step: a Claude Code / Gemini CLI / Cursor hook, a Claude Code status line or credential helper command, or a VS Code task that runs when the folder opens. Opening the project in that tool runs it with the developer's privileges. An `http` hook (here or in a plugin or skill) posts each event it is bound to, tool input and output included, to a server that is not this machine.
"command": "python3 .claude/statusline.py"
Fix: Read the command and anything it runs, or the URL a hook posts to, before opening this project in an agent or editor. A published package has no reason to ship project auto-run config.
The component reads files from disk.
const packageJsonContent = fs.readFileSync(packageJsonPath, 'utf-8');
let content = fs.readFileSync(mainCjsPath, 'utf8');
const packageJson = JSON.parse(fs.readFileSync(packageJsonPath, 'utf8'));
const mainContent = fs.readFileSync(mainCjsPath, 'utf8');
return await fs.readFile(filePath, 'utf-8');
content = await fs.readFile(filePath, 'utf-8');
const existingContent = await fs.readFile(fullPath, 'utf-8');
const existingContent = await fs.readFile(fullPath, 'utf-8');
currentContent = await fs.readFile(fullPath, 'utf-8');
const existingContent = await fs.readFile(fullPath, 'utf-8');
currentContent = await fs.readFile(fullPath, 'utf-8');
const actual = await fs.readFile(fullPath);
const actual = await fs.readFile(fullPath, 'utf-8');
localContent = await fs.readFile(fullPath, 'utf-8');
return await fs.readFile(filePath, 'utf-8');
const buffer = await fs.readFile(filePath);
localContent = await fs.readFile(fullPath, 'utf-8');
const content = fs.readFileSync(path.join(dir, 'packmind.json'), 'utf-8');
const content = fs.readFileSync(lockFilePath, 'utf-8');
cas.push(fs.readFileSync(certFile));
const configContent = await fs.readFile(configPath, 'utf-8');
return await fs.readFile(filePath, 'utf-8');
const content = await fs.readFile(lockFilePath, 'utf-8');
const content = fs.readFileSync(this.storagePath, 'utf-8');
const content = fs.readFileSync(credentialsPath, 'utf-8');
Why it matters: Usually legitimate, but worth confirming it can't be steered into reading sensitive files.
Fix: Confirm which files are read and that paths cannot be influenced by untrusted input to reach sensitive locations.
The component writes or deletes files on disk.
fs.writeFileSync(mainCjsPath, content);
fs.writeFileSync(packageJsonPath, JSON.stringify(packageJson, null, 2) + '\n');
fs.writeFileSync(mainCjsPath, '#!/usr/bin/env node\n' + mainContent);
fs.writeFileSync(path.join(moduleDir, 'index.js'), stubContent.trim());
fs.writeFileSync(
await fs.writeFile(fullPath, file.content, 'utf-8');
await fs.writeFile(fullPath, file.content, 'utf-8');
await fs.writeFile(fullPath, buffer);
await fs.writeFile(fullPath, finalContent, 'utf-8');
await fs.writeFile(fullPath, content, 'utf-8');
await fs.unlink(fullPath);
await fs.writeFile(fullPath, mergedContent, 'utf-8');
await fs.rm(fullPath, { recursive: true, force: true });await fs.unlink(fullPath);
await fs.rm(fullPath, { recursive: true, force: true });await fs.rmdir(currentDir);
await fs.writeFile(fullPath, buffer);
await fs.writeFile(fullPath, finalContent, 'utf-8');
await fs.writeFile(fullPath, content, 'utf-8');
await fs.unlink(fullPath);
await fs.writeFile(fullPath, mergedContent, 'utf-8');
await fs.rm(fullPath, { recursive: true, force: true });await fs.unlink(fullPath);
await fs.rm(fullPath, { recursive: true, force: true });await fs.rmdir(currentDir);
Why it matters: Usually legitimate, but worth confirming the paths can't be controlled by untrusted input.
Fix: Confirm which files are written/deleted and that paths cannot be influenced by untrusted input.
The component reads files from disk.
result = parse(path.read_text(encoding="utf-8"))
text = path.read_text(encoding="utf-8")
for match in TASK_STATE.finditer(plan_path.read_text(encoding="utf-8")):
content = path.read_text(encoding="utf-8")
result = parse(path.read_text(encoding="utf-8"))
text = path.read_text(encoding="utf-8")
for match in TASK_STATE.finditer(plan_path.read_text(encoding="utf-8")):
content = path.read_text(encoding="utf-8")
result = parse(path.read_text(encoding="utf-8"))
text = path.read_text(encoding="utf-8")
for match in TASK_STATE.finditer(plan_path.read_text(encoding="utf-8")):
content = path.read_text(encoding="utf-8")
result = parse(path.read_text(encoding="utf-8"))
text = path.read_text(encoding="utf-8")
for match in TASK_STATE.finditer(plan_path.read_text(encoding="utf-8")):
content = path.read_text(encoding="utf-8")
result = parse(path.read_text(encoding="utf-8"))
text = path.read_text(encoding="utf-8")
for match in TASK_STATE.finditer(plan_path.read_text(encoding="utf-8")):
content = path.read_text(encoding="utf-8")
content = skill_md.read_text()
result = parse(path.read_text(encoding="utf-8"))
text = path.read_text(encoding="utf-8")
for match in TASK_STATE.finditer(plan_path.read_text(encoding="utf-8")):
content = path.read_text(encoding="utf-8")
Why it matters: Usually legitimate, but worth confirming it can't be steered into reading sensitive files.
Fix: Confirm which files are read and that paths cannot be influenced by untrusted input to reach sensitive locations.
The component writes or deletes files on disk.
path.write_text(content, encoding="utf-8")
path.write_text(content, encoding="utf-8")
path.write_text(content, encoding="utf-8")
path.write_text(content, encoding="utf-8")
path.write_text(content, encoding="utf-8")
skill_md_path.write_text(skill_content)
example_script.write_text(EXAMPLE_SCRIPT.format(skill_name=skill_name))
example_reference.write_text(EXAMPLE_REFERENCE.format(skill_title=skill_title))
example_asset.write_text(EXAMPLE_ASSET)
path.write_text(content, encoding="utf-8")
Why it matters: Usually legitimate, but worth confirming the paths can't be controlled by untrusted input.
Fix: Confirm which files are written/deleted and that paths cannot be influenced by untrusted input.
package.json has a 'prepare' script (runs on git/local installs and before publishing).
"prepare": "node scripts/prepare.mjs",
Why it matters: Usually a build step, but confirm it doesn't fetch or run remote code.
Fix: Usually a legitimate build step; confirm it only builds and does not fetch or execute remote code.
The component makes outbound network requests.
import axios from 'axios';
const response = await axios.post<typeof conversionData>(
if (axios.isAxiosError(error)) {import axios from 'axios';
await axios.post(WEBHOOK_URL, payload, {nodemailer: STUB_MODULE,
'nodemailer',
'@types/nodemailer',
'nodemailer',
'@types/nodemailer',
'nodemailer',
'@types/nodemailer',
import * as http from 'http';
const response = await fetch(url, {const response = await fetch(url, {const response = await fetch(`${host}/api/v0/auth/me`, {const response = await fetch(
const response = await fetch(
await fetch("/api/invoices", { method: "POST", body: JSON.stringify(invoice) })import axios from "axios" // ❌ domain depends on infrastructure
await axios.post("/users", this)import axios, { AxiosInstance, AxiosRequestConfig } from 'axios';this._axiosInstance = axios.create({const response = await fetch(`${apiUrl}/v0/sse/subscribe`, {const response = await fetch(`${apiUrl}/v0/sse/unsubscribe`, {Why it matters: Usually legitimate, but confirm the destinations are expected and no sensitive data leaves.
Fix: Confirm the destination hosts are expected and that no sensitive data is sent off-host.
An MCP tool surface (manifest or tool definitions) was found.
"mcpServers": {Why it matters: Just context — review which tools it offers and their permissions.
Fix: Review the declared MCP tools and their permissions.
How attackers abuse these capabilities
Interactive labs on the attack class behind the rules above. They show the technique, not anything found in Packmind MCP server.
The free Claude Code plugin scans each npm, PyPI or MCP package with this report's open-source engine before the install command runs. It runs on your machine and needs no account.
Check your own component
Run the same evidence-backed scan on any MCP server, agent skill, or package.
Scan your own componentHow we determine this: deterministic static analysis (regex + AST), evidence-anchored, no code execution. Methodology →