Postman installed itself into Codex and Claude
Opening Postman on October 24 installed its plugin into two coding agents the user never asked it to touch — Codex and Claude Code — and the installer's own code then wrote the trust record for its own startup hook. No API data was shown to leave the machine: the remote MCP connection failed authentication, and the reporter explicitly does not claim exfiltration. That correction matters, because…
On October 24, 2026, Postman's app-support team installed its plugin into two coding agents, Codex and Claude Code, without the user's consent. The installer then wrote its own trust record for its startup hook, bypassing the user's approval. This occurred silently, without any API data being shown to leave the machine, though the remote MCP connection failed authentication.
The headline of the story, which suggested that Postman had shared all API data, is inaccurate. The issue is documented in a GitHub issue filed against Postman's app-support tracker, which provides detailed logs from the incident. The three key claims are that the startup hook reroutes broad API work, the installer approves its own hook, and the remote MCP endpoint fails to connect.
While the issue was not reproducible on a clean Mac, it highlights the need for explicit user consent for these actions. The recommended fix is to prevent automatic agent-plugin installation for existing desktop users and to narrow the startup instruction to explicit Postman requests.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
- Anthropic speeds up Claude Code with message threads. But your token bill stays the same. thenewstack.io
- I measured 8 weeks of Claude Code: 87% lower cost, 86% first-try typecheck dev.to
- I made Claude Code, Codex, and Antigravity build their own rivals, and one was in a different league xda-developers.com
- Is Claude Code truly faster with message threads? I did the math on my logs zdnet.com
- Qodo builds context into AI code review with Codex openai.com
- to experiment with Claude’s new Code Interpreter substack.com
