Historial de auditorías

references/tunnel/patterns.md:184

Pipe to shell pattern

await sandbox.exec('curl -fsSL https://code-server.dev/install.sh | sh');

The reference contains a pipe-to-shell install command that an agent or user could copy into a live environment. Even though it is documentation, this pattern is hazardous and should be replaced with safer installation steps.

Alto

Systemd service enablement

systemctl start cloudflared && systemctl enable cloudflared

The reference includes privileged service-management commands. They are documentation examples, but they can alter host persistence or privileges if followed without review.

Elementos de revisión de capacidades (31)

Estas son capacidades locales reales que pueden esperarse para esta habilidad, por lo que requieren revisión, pero no se cuentan como comportamiento malicioso confirmado.

Alto

references/flagship/README.md:31

Dynamic code evaluation with eval()

| **OpenFeature SDK** | Workers, Node.js, browser | `@cloudflare/flagship` + `@openfeature/server-sd

The reference mentions dynamic eval usage, which is dangerous if copied into application code. Context may be educational, so confidence is moderate rather than high.

Alto

Process exec

const cursor = this.sql.exec('SELECT * FROM users WHERE email = ?', email);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec('UPDATE accounts SET balance = balance - ? WHERE id = ?', 100, 1);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

references/do-storage/configuration.md:44-49 references/do-storage/patterns.md:13-16

Process exec

this.sql.exec('UPDATE accounts SET balance = balance + ? WHERE id = ?', 100, 2);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec · 2 apariciones

this.sql.exec(`

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec("CREATE TABLE events(id INTEGER PRIMARY KEY)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec("INSERT INTO events VALUES (?)", 1234567890123456789n); // Corrupts!

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec("CREATE TABLE events(id TEXT PRIMARY KEY)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec("INSERT INTO events VALUES (?)", "1234567890123456789");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec(`CREATE TABLE users(id INTEGER PRIMARY KEY, name TEXT)`);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec("INSERT INTO _sql_schema_migrations (id) VALUES (1)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec(`ALTER TABLE users ADD COLUMN email TEXT`);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec("INSERT INTO _sql_schema_migrations (id) VALUES (2)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec('DELETE FROM requests WHERE key = ? AND timestamp < ?', key, now - window);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

const cursor = this.sql.exec('SELECT * FROM users WHERE email = ?', email);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec('UPDATE accounts SET balance = balance - ? WHERE id = ?', 100, 1);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

references/do-storage/configuration.md:44 references/do-storage/patterns.md:13

Python exec() function

this.sql.exec('UPDATE accounts SET balance = balance + ? WHERE id = ?', 100, 2);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function · 2 apariciones

this.sql.exec(`

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec("CREATE TABLE events(id INTEGER PRIMARY KEY)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec("INSERT INTO events VALUES (?)", 1234567890123456789n); // Corrupts!

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec("CREATE TABLE events(id TEXT PRIMARY KEY)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec("INSERT INTO events VALUES (?)", "1234567890123456789");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec(`CREATE TABLE users(id INTEGER PRIMARY KEY, name TEXT)`);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec("INSERT INTO _sql_schema_migrations (id) VALUES (1)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec(`ALTER TABLE users ADD COLUMN email TEXT`);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec("INSERT INTO _sql_schema_migrations (id) VALUES (2)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

references/tunnel/patterns.md:190

Python exec() function

this.sql.exec('DELETE FROM requests WHERE key = ? AND timestamp < ?', key, now - window);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

sudo privilege escalation

sudo cloudflared service install

The reference includes privileged service-management commands. They are documentation examples, but they can alter host persistence or privileges if followed without review.

Alto

references/tunnel/patterns.md:191

sudo privilege escalation

sudo launchctl start com.cloudflare.cloudflared

The reference includes privileged service-management commands. They are documentation examples, but they can alter host persistence or privileges if followed without review.

Factores de riesgo

⚙️ Comandos externos (50)

🌐 Acceso a red (50)

🔑 Variables de entorno (50)

📁 Acceso al sistema de archivos (50)

⚡ Contiene scripts (50)

Patrones detectados

Pipe to shell patternSystemd service enablement

Auditado por: claude

Versión de auditoría 4

9 jul 2026, 01:30

321

Archivos escaneados

49,016

Líneas analizadas

Elementos de revisión

Falsos positivos ignorados

Preocupaciones de seguridad confirmadas (2)

Crítico

references/tunnel/patterns.md:184

Pipe to shell pattern

await sandbox.exec('curl -fsSL https://code-server.dev/install.sh | sh');

Alto

Systemd service enablement

systemctl start cloudflared && systemctl enable cloudflared

The reference includes privileged service-management commands. They are documentation examples, but they can alter host persistence or privileges if followed without review.

Elementos de revisión de capacidades (31)

Estas son capacidades locales reales que pueden esperarse para esta habilidad, por lo que requieren revisión, pero no se cuentan como comportamiento malicioso confirmado.

Alto

references/flagship/README.md:31

Dynamic code evaluation with eval()

| **OpenFeature SDK** | Workers, Node.js, browser | `@cloudflare/flagship` + `@openfeature/server-sd

The reference mentions dynamic eval usage, which is dangerous if copied into application code. Context may be educational, so confidence is moderate rather than high.

Alto

Process exec

const cursor = this.sql.exec('SELECT * FROM users WHERE email = ?', email);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec('UPDATE accounts SET balance = balance - ? WHERE id = ?', 100, 1);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

references/do-storage/configuration.md:44-49 references/do-storage/patterns.md:13-16

Process exec

this.sql.exec('UPDATE accounts SET balance = balance + ? WHERE id = ?', 100, 2);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec · 2 apariciones

this.sql.exec(`

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec("CREATE TABLE events(id INTEGER PRIMARY KEY)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec("INSERT INTO events VALUES (?)", 1234567890123456789n); // Corrupts!

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec("CREATE TABLE events(id TEXT PRIMARY KEY)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec("INSERT INTO events VALUES (?)", "1234567890123456789");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec(`CREATE TABLE users(id INTEGER PRIMARY KEY, name TEXT)`);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec("INSERT INTO _sql_schema_migrations (id) VALUES (1)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec(`ALTER TABLE users ADD COLUMN email TEXT`);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec("INSERT INTO _sql_schema_migrations (id) VALUES (2)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Process exec

this.sql.exec('DELETE FROM requests WHERE key = ? AND timestamp < ?', key, now - window);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

const cursor = this.sql.exec('SELECT * FROM users WHERE email = ?', email);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec('UPDATE accounts SET balance = balance - ? WHERE id = ?', 100, 1);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

references/do-storage/configuration.md:44 references/do-storage/patterns.md:13

Python exec() function

this.sql.exec('UPDATE accounts SET balance = balance + ? WHERE id = ?', 100, 2);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function · 2 apariciones

this.sql.exec(`

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec("CREATE TABLE events(id INTEGER PRIMARY KEY)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec("INSERT INTO events VALUES (?)", 1234567890123456789n); // Corrupts!

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec("CREATE TABLE events(id TEXT PRIMARY KEY)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec("INSERT INTO events VALUES (?)", "1234567890123456789");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec(`CREATE TABLE users(id INTEGER PRIMARY KEY, name TEXT)`);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec("INSERT INTO _sql_schema_migrations (id) VALUES (1)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec(`ALTER TABLE users ADD COLUMN email TEXT`);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

Python exec() function

this.sql.exec("INSERT INTO _sql_schema_migrations (id) VALUES (2)");

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

references/tunnel/patterns.md:190

Python exec() function

this.sql.exec('DELETE FROM requests WHERE key = ? AND timestamp < ?', key, now - window);

The reference includes process or dynamic execution patterns. They appear to be examples, but such patterns are security-sensitive and should stay flagged for manual review.

Alto

sudo privilege escalation

sudo cloudflared service install

The reference includes privileged service-management commands. They are documentation examples, but they can alter host persistence or privileges if followed without review.

Alto

references/tunnel/patterns.md:191

sudo privilege escalation

sudo launchctl start com.cloudflare.cloudflared

The reference includes privileged service-management commands. They are documentation examples, but they can alter host persistence or privileges if followed without review.

Factores de riesgo

⚙️ Comandos externos (50)

🌐 Acceso a red (50)

🔑 Variables de entorno (50)

📁 Acceso al sistema de archivos (50)

⚡ Contiene scripts (50)

Patrones detectados

Pipe to shell patternSystemd service enablement

Auditado por: claude

Versión de auditoría 3

5 jul 2026, 07:16

Most static findings are false positives from Markdown, Cloudflare API examples, SQL exec calls, and documented placeholder secrets. Confirmed risks remain in Sandbox examples that promote unsafe shell execution, token-in-URL cloning, and untrusted command construction, plus Tunnel service-install examples that require administrator approval.

321

Archivos escaneados

49,016

Líneas analizadas

Elementos de revisión

Falsos positivos ignorados

Preocupaciones de seguridad confirmadas (4)

Crítico

references/tunnel/patterns.md:184

Pipe to shell pattern

await sandbox.exec('curl -fsSL https://code-server.dev/install.sh | sh');

The example pipes a remotely fetched install script directly into sh inside a sandbox. This is a classic unsafe installation pattern and should not be promoted as a common pattern.

Alto

Systemd service enablement

systemctl start cloudflared && systemctl enable cloudflared

The command enables a persistent systemd service. This is legitimate for Tunnel setup but creates persistence and should be clearly gated behind admin approval.

Alto

multiple:1

[HEURISTIC] SUSPICIOUS COMBINATION: Code execution + Persistence mechanism

This combination is common in malware that maintains access

The repository contains confirmed pipe-to-shell and persistent service-install examples. The evidence is documentation rather than hidden executable code, so confidence is moderate.

Alto

references/sandbox/patterns.md:39-41 references/sandbox/patterns.md:135-138 references/sandbox/patterns.md:185-186

Unsafe Sandbox Command Patterns Presented as Examples

The Sandbox reference includes common-pattern examples that pipe a remote installer into a shell, interpolate request values into git commands, and execute request body code. These examples need stronger warnings and safer alternatives.

The cited lines present these commands under common patterns, not only as anti-patterns. Static analysis finds the mechanics, but the framing as recommended examples is the additional risk.

Elementos de revisión de capacidades (23)

Estas son capacidades locales reales que pueden esperarse para esta habilidad, por lo que requieren revisión, pero no se cuentan como comportamiento malicioso confirmado.

Crítico

Python exec() function

await sandbox.exec('curl -fsSL https://code-server.dev/install.sh | sh');

This finding is tied to the same curl-to-shell sandbox example. It fetches code from the network and executes it without pinning or verification.

Crítico

Process exec

await sandbox.exec('curl -fsSL https://code-server.dev/install.sh | sh');

This finding is tied to the same curl-to-shell sandbox example. It fetches code from the network and executes it without pinning or verification.

Crítico

references/sandbox/api.md:167

Hardcoded URL

await sandbox.exec('curl -fsSL https://code-server.dev/install.sh | sh');

This finding is tied to the same curl-to-shell sandbox example. It fetches code from the network and executes it without pinning or verification.

Alto

Python exec() function

const result = await sandbox.exec('python /tmp/code.py');

The lifecycle example writes a variable named code to /tmp/code.py and executes it. This is safe only for trusted code and needs explicit validation guidance.

Alto

references/sandbox/api.md:167

Process exec

const result = await sandbox.exec('python /tmp/code.py');

The lifecycle example writes a variable named code to /tmp/code.py and executes it. This is safe only for trusted code and needs explicit validation guidance.

Alto

references/sandbox/api.md:166

Temp directory access

await sandbox.writeFile('/tmp/code.py', code);

The lifecycle example writes a variable named code to /tmp/code.py and executes it. This is safe only for trusted code and needs explicit validation guidance.

Alto

references/sandbox/api.md:167

Temp directory access

const result = await sandbox.exec('python /tmp/code.py');

The lifecycle example writes a variable named code to /tmp/code.py and executes it. This is safe only for trusted code and needs explicit validation guidance.

Alto

references/sandbox/gotchas.md:29

Python exec() function

return await sandbox.exec(cmd);

The retry helper executes a caller-supplied command string. It may be legitimate, but it needs a validation boundary to avoid command injection.

Alto

references/sandbox/gotchas.md:29

Process exec

return await sandbox.exec(cmd);

The retry helper executes a caller-supplied command string. It may be legitimate, but it needs a validation boundary to avoid command injection.

Alto

references/sandbox/patterns.md:138

Python exec() function

await sandbox.exec(`git clone -b ${branch} ${repo} /workspace/repo`);

The CI example interpolates request-supplied repo and branch values directly into a shell command. This creates a command-injection risk unless inputs are strictly validated.

Alto

references/sandbox/patterns.md:186

Python exec() function

const result = await session.exec(`python3 -c "${code}"`);

The session example executes request body content through python3 -c with shell interpolation. This is unsafe untrusted code execution guidance.

Alto

references/sandbox/patterns.md:138

Python exec() function

await sandbox.exec(`git clone https://${env.GITHUB_TOKEN}@github.com/user/repo.git`);

The example embeds GITHUB_TOKEN directly in a git clone URL. Tokens in URLs can leak through logs, shell history, process listings, and error messages.

Alto

Process exec

await sandbox.exec(`git clone -b ${branch} ${repo} /workspace/repo`);

The CI example interpolates request-supplied repo and branch values directly into a shell command. This creates a command-injection risk unless inputs are strictly validated.

Alto

references/sandbox/patterns.md:186

Process exec

const result = await session.exec(`python3 -c "${code}"`);

The session example executes request body content through python3 -c with shell interpolation. This is unsafe untrusted code execution guidance.

Alto

references/sandbox/patterns.md:138-159

Process exec

await sandbox.exec(`git clone https://${env.GITHUB_TOKEN}@github.com/user/repo.git`);

The example embeds GITHUB_TOKEN directly in a git clone URL. Tokens in URLs can leak through logs, shell history, process listings, and error messages.

Alto

Ruby/shell backtick execution

await sandbox.exec(`git clone -b ${branch} ${repo} /workspace/repo`);

The CI example interpolates request-supplied repo and branch values directly into a shell command. This creates a command-injection risk unless inputs are strictly validated.

Alto

references/sandbox/patterns.md:186-191

Ruby/shell backtick execution

const result = await session.exec(`python3 -c "${code}"`);

The session example executes request body content through python3 -c with shell interpolation. This is unsafe untrusted code execution guidance.

Alto

references/sandbox/patterns.md:200-201

Ruby/shell backtick execution

await sandbox.exec(`git clone https://${env.GITHUB_TOKEN}@github.com/user/repo.git`);

The example embeds GITHUB_TOKEN directly in a git clone URL. Tokens in URLs can leak through logs, shell history, process listings, and error messages.

Alto

Hardcoded URL

await sandbox.exec(`git clone https://${env.GITHUB_TOKEN}@github.com/user/repo.git`);

The example embeds GITHUB_TOKEN directly in a git clone URL. Tokens in URLs can leak through logs, shell history, process listings, and error messages.

Alto