CVE-2026-77938

Letta through 0.16.8 contains an information disclosure vulnerability that allows users running tools to read server secrets because LocalSandbox subprocesses inherit the full server environment. Attackers can run code via POST /v1/tools/run or custom tools to read LETTAPGURI, POSTGRESPASSWORD, a...

CVE-2026-77936

Letta through 0.16.8 ships a compose.yaml that does not enable LETTASERVERSECURE, so the password middleware is never added and all API routes are unauthenticated, and it exposes a superuser PostgreSQL port with default credentials. Remote attackers can execute arbitrary Python via POST...

CVE-2026-77938 Letta through 0.16.8 Credential Exposure via Tool Sandbox Environment Inheritance

Letta through 0.16.8 contains an information disclosure vulnerability that allows users running tools to read server secrets because LocalSandbox subprocesses inherit the full server environment. Attackers can run code via POST /v1/tools/run or custom tools to read LETTAPGURI, POSTGRESPASSWORD, a...

CVE-2026-77938 Letta through 0.16.8 Credential Exposure via Tool Sandbox Environment Inheritance

Letta through 0.16.8 contains an information disclosure vulnerability that allows users running tools to read server secrets because LocalSandbox subprocesses inherit the full server environment. Attackers can run code via POST /v1/tools/run or custom tools to read LETTAPGURI, POSTGRESPASSWORD, a...

CVE-2026-77938

Letta versions through 0.16.8 are vulnerable to an information disclosure vulnerability due to LocalSandbox subprocesses inheriting the full server environment. An attacker can execute code via POST /v1/tools/run or custom tools to extract sensitive secrets, including LETTA_PG_URI , POSTGRES_PASS...

EUVD-2026-104694

Letta through 0.16.8 contains an information disclosure vulnerability that allows users running tools to read server secrets because LocalSandbox subprocesses inherit the full server environment. Attackers can run code via POST /v1/tools/run or custom tools to read LETTAPGURI, POSTGRESPASSWORD, a...

CVE-2026-77938

Letta through 0.16.8 contains an information disclosure vulnerability that allows users running tools to read server secrets because LocalSandbox subprocesses inherit the full server environment. Attackers can run code via POST /v1/tools/run or custom tools to read LETTAPGURI, POSTGRESPASSWORD, a...

EUVD-2026-104692

Letta through 0.16.8 ships a compose.yaml that does not enable LETTASERVERSECURE, so the password middleware is never added and all API routes are unauthenticated, and it exposes a superuser PostgreSQL port with default credentials. Remote attackers can execute arbitrary Python via POST...

CVE-2026-77936

Letta through 0.16.8 ships a compose.yaml that does not enable LETTASERVERSECURE, so the password middleware is never added and all API routes are unauthenticated, and it exposes a superuser PostgreSQL port with default credentials. Remote attackers can execute arbitrary Python via POST...

PT-2026-117956

Name of the Vulnerable Software and Affected Versions Letta versions prior to 0.16.9 Description An information disclosure issue exists where LocalSandbox subprocesses inherit the full server environment. This allows users to read server secrets by executing code through the 'POST /v1/tools/run'...

Letta 0.7.12 - Remote Code Execution

Unauthenticated RCE via POST /v1/tools/run executing attacker tool sourcecode. !/usr/bin/env python3 -- coding: utf-8 -- """Letta 0.7.12 - Remote Code Execution""" import hashlib import json from kittysploit import from lib.protocols.http.httpclient import Httpclient class ModuleScanner,...

CVE-2025-51482

Remote Code Execution in letta.server.restapi.routers.v1.tools.runtoolfromsource in letta-ai Letta 0.7.12 allows remote attackers to execute arbitrary Python code and system commands via crafted payloads to the /v1/tools/run endpoint, bypassing intended sandbox restrictions...

CVE-2025-51482: Improper Control of Generation of Code

Remote Code Execution in letta.server.restapi.routers.v1.tools.runtoolfromsource in letta-ai Letta 0.7.12 allows remote attackers to execute arbitrary Python code and system commands via crafted payloads to the /v1/tools/run endpoint, bypassing intended sandbox restrictions...