CVE-2026-100867

Versions of spaceship-prompt up to and including 4.22.5 are vulnerable to Terminal Escape Sequence Injection . The software fails to sanitize control characters within project manifest version fields before rendering them in the zsh prompt . An attacker can embed ANSI/OSC escape sequences in thes...

EUVD-2026-87939

spaceship-prompt through 4.22.5 fails to sanitize control characters from project manifest version fields before rendering them in the zsh prompt. Attackers can embed ANSI/OSC escape sequences in version fields of package manifests to manipulate terminal output, rewrite window titles, or spoof...

CVE-2026-100867 spaceship-prompt through 4.22.5 Terminal Escape Sequence Injection

spaceship-prompt through 4.22.5 fails to sanitize control characters from project manifest version fields before rendering them in the zsh prompt. Attackers can embed ANSI/OSC escape sequences in version fields of package manifests to manipulate terminal output, rewrite window titles, or spoof...

Can Prompt Anonymity Protect Your Identity from LLM Providers?

User conversations with large language models LLMs often contain highly sensitive personal information that can be exploited by LLM providers to create detailed user dossiers, enable targeted advertising, and train more powerful models. To protect user privacy, anonymizing LLM proxies have emerge...

Climbing the Hill: Prompt Injection Red-Teaming against Frontier Models with Curriculum Reinforcement Learning

Prompt injection is a leading security risk for LLMs and LLM-based applications such as agents. State-of-the-art red-teaming methods for prompt injection leverage reinforcement learning RL to train an attacker LLM to generate effective injected prompts. However, when targeting frontier LLMs such ...

CVE-2026-100867: Improper Neutralization of Escape, Meta, or Control Sequences

spaceship-prompt through 4.22.5 fails to sanitize control characters from project manifest version fields before rendering them in the zsh prompt. Attackers can embed ANSI/OSC escape sequences in version fields of package manifests to manipulate terminal output, rewrite window titles, or spoof...

PT-2026-99537

Name of the Vulnerable Software and Affected Versions spaceship-prompt versions prior to 4.22.6 Description The software fails to sanitize control characters from project manifest version fields before rendering them in the zsh prompt. This allows attackers to embed ANSI/OSC escape...

API Secrets Should Never Become Tokens in the LLM's Vocabulary

Tool-using large language model LLM agents turn credential hygiene from a storage problem into an execution-security problem. A key pasted into a prompt, or embedded in a system prompt or tool configuration, crosses from an authentication boundary into a data pipeline, where it may persist in...

When Consent Outlives Context: Residual Authority Replay in Long-Lived Agents

LLM agents increasingly rely on user approval to authorize security-sensitive actions at runtime. Such approvals are granted within a specific task and execution context. In long-lived agents, authorization decisions may need to persist across tasks or sessions. We find that this continuity can...

CVE-2026-100651

A flaw was found in vLLM. The service fails to properly enforce length limits on input prompts when handling certain multimodal inference requests. A remote attacker with access to the inference service could exploit this vulnerability by submitting an excessively long prompt, causing an internal...

EUVD-2026-87745

vLLM before 0.29.0 fails to enforce decoder prompt-length validation on the disaggregated serving endpoint /inference/v1/generate. When the request contains a 'features' multimodal payload, vllm/entrypoints/serve/disagg/serving.py builds a multimodal EngineInput directly from the caller-supplied...

CVE-2026-100651 vllm before 0.29.0 Denial of Service via Decoder Prompt Length Bypass

vLLM before 0.29.0 fails to enforce decoder prompt-length validation on the disaggregated serving endpoint /inference/v1/generate. When the request contains a 'features' multimodal payload, vllm/entrypoints/serve/disagg/serving.py builds a multimodal EngineInput directly from the caller-supplied...

CVE-2026-100585

OpenClaw npm package openclaw before 2026.7.1 fails to enforce the owner-only authorization requirement for Claude Code permission prompts delivered through the MCP channel bridge. An authorized non-owner channel sender with channel command access can approve or deny a pending permission request...

CVE-2026-100585 OpenClaw before 2026.7.1 Authentication Bypass via MCP Channel

OpenClaw npm package openclaw before 2026.7.1 fails to enforce the owner-only authorization requirement for Claude Code permission prompts delivered through the MCP channel bridge. An authorized non-owner channel sender with channel command access can approve or deny a pending permission request...

CVE-2026-100585

OpenClaw npm package "openclaw" before 2026.7.1 fails to enforce the owner-only authorization requirement for Claude Code permission prompts delivered through the MCP channel bridge. An authorized non-owner channel sender with channel command access can approve or deny a pending permission reques...

CVE-2026-100585

The OpenClaw npm package (openclaw) versions prior to 2026.7.1 contains an authorization bypass vulnerability. The software fails to enforce an owner-only requirement for Claude Code permission prompts sent via the MCP channel bridge . This allows an authorized non-owner channel sender with comma...

EUVD-2025-26607

Incomplete validation of dunder attributes allows an attacker to escape from the Local Python execution environment sandbox, enforced by smolagents. The attack requires a Prompt Injection in order to trick the agent to create malicious code...

PT-2026-99322

Name of the Vulnerable Software and Affected Versions vLLM versions prior to 0.29.0 Description The software fails to enforce decoder prompt-length validation on the disaggregated serving endpoint '/inference/v1/generate'. When a request includes a 'features' multimodal payload, the system builds...

PT-2026-99221

Name of the Vulnerable Software and Affected Versions OpenClaw versions prior to 2026.7.1 Description The openclaw npm package fails to enforce owner-only authorization for Claude Code permission prompts sent via the MCP channel bridge. This allows an authorized non-owner with channel command...

Configuration, Not Conscience: A Large-Scale Empirical Study of LLM System Prompts

Leaked system prompts are often treated as windows into the hidden values of commercial language models, yet their composition is rarely studied at scale. We analyze a merged corpus of 407 leaked, reconstructed, or officially published system prompts from 62 vendors across four community...