Fetching from the wire…
01
02
03
04
05
06
07
08
Security2026-06-24 · source-backed
The Cloud Security Alliance research note generalizes a rule you should tattoo on your wall. Anything an agent pulls through an MCP tool (error reports, tickets, search results, web pages) is untrusted input, not trusted system output. An attacker who can write into a Sentry error or a Jira ticket can plant instructions your agent will execute. The defense is boring and correct: never let an agent act on instructions embedded in tool output, scope tool credentials tightly, gate irreversible actions behind explicit approval.
Each link below shares sources, entities, or timing with this story.
Claude Code uses MCP / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Claude Code uses MCP); both cover Agentjacking, Cloud Security Alliance, MCP, Sentry; overlapping topics (agent, agentjacking, error, input).
Cursor uses MCP / Shared entities / Shared topic / What happened next
Linked by a graph relationship (Cursor uses MCP); both cover Agentjacking, MCP, Sentry; overlapping topics (agent, agentjacking).
Claude uses MCP / Shared entities / Shared topic / What happened next
Linked by a graph relationship (Claude uses MCP); both cover MCP, Sentry; overlapping topics (agent, instruction, tool, untrusted).
Claude Code uses MCP / Shared entities / Shared topic / Earlier coverage / Tension
Linked by a graph relationship (Claude Code uses MCP); both cover Jira, MCP; overlapping topics (agent, tool).
Claude Code uses MCP / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Claude Code uses MCP); both cover MCP, Sentry; overlapping topics (action, agent, approval).
Claude Code uses MCP / Shared entity: MCP / Shared topic / What happened next / Tension
Linked by a graph relationship (Claude Code uses MCP); both cover MCP; overlapping topics (agent, instruction, tool).
OpenAI supports MCP / Shared entity: MCP / Shared topic / What happened next / Tension
Linked by a graph relationship (OpenAI supports MCP); both cover MCP; overlapping topics (agent, instruction, tool).
Anthropic released MCP / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Anthropic released MCP); both cover MCP, Sentry; overlapping topics (agent, credential).