En-tête négligent

Laboratoires de piratage

WorkOS ARTEX incident: AI agent scope and API exposure, illustrated by glass trust boundaries.

WorkOS ARTEX Incident: AI Agent Scope and API Exposure

Three independent audit rows cover package source, persistent instruction differences and effective provider routing settings.

Prompt Adware: Audit AI Coding Tool Instructions and Cleanup

A threat brief is mapped to an owned asset, then to separate control and runtime evidence checks; reported activity does not establish compromise.

Cloudflare Zero Wire: Turn Vulnerability Intelligence into Asset Evidence

Source revision, build artifact and runtime deployment converge on a review classified as validated, not reproduced or inconclusive.

Arnica Agentic Pen Testing: Verify SAST Findings Without Losing Runtime Context

A layered verification path follows header construction to an HTTP/1.1 upstream and request boundary evidence; injection alone does not prove desync.

CRLF-Powered Desync: Verify HTTP Request Header Injection Safely

An exposed secret maps to an application identity and issued tokens, with separate actions to revoke the secret, review access and check token expiry.

ARTEX and Exposed Secrets: Revocation and Evidence Checks

A conceptual review loop connects a test scenario, a release version and its result evidence.

Escape Continuous AI Pentesting: Evaluate Coverage, Ownership and Retest Evidence

Two independent checks: authorization for users calling an app and the allowed outbound source for app calls to Shopify Admin API.

Shopify App IP Allowlists: Serverless Setup, Deadlines and 403 Verification

Inspector 2.14.0 fixes the app frame, input handling and schema issues. Legacy 1.0.4 addresses the schema-validation stall among these three advisories.

MCP Inspector Security Update: What 2.14.0 Fixes and How to Verify Your Setup

MCP server pre-connection audit covering publisher source, maintenance lifecycle, pinned version and minimum agent permissions.

MCPhantom and Deprecated MCP Servers: What to Audit Before Connecting an AI Agent

Pied de page négligent

French