MCP security researcher: killing MCP means you don't care about security, CLIs lack guarantees
yenkel · x · 2026-09-16
Amid the "MCP is dead, CLIs are better" debate, MCP security researcher yenkel resurfaces his article arguing that CLI-to-API calls don't provide the guarantees needed for good security, while code mode against an API (no CLI) can — but only if the API is sufficiently close to spec. The quoted post counters that MCP beats CLIs for most integrations: models are far better at tool calling, deferred tools work, MCP is now stateless, and adding params like query handles data composition/filtering.
Related event: Developers Flip Toward MCP Over CLI for Agent Tool Integration(5 posts)→
More from coding & agent
- Astra x GPT-Images 2.5: agentic perception meets image consistency — soumitrashukla9 · 2026-09-16
- Dev builds Sturzi, a Grok bot that finds and files simpler versions of Argentine regulations — StewartalsopIII · 2026-09-16
- Compaction vs prompt caching: subagent context isolation may beat summarizing in production agents — Marcus_MSC · 2026-09-16
- Sakana Marlin adds Interactive Reading and PowerPoint export for deep research reports — SakanaAILabs · 2026-09-16
- Andy Matuschak demos programmable highlighters that trigger AI actions on annotations — andy_matuschak · 2026-09-16
- Dev shares trick: dedicated subagent filters SSH output to protect agent context — SaleemUsama · 2026-09-16