AI article

What an Empty Prompt Costs: Measure the Baseline Tokens of 4 Coding Agent CLIs

Community description: Every fresh agent session sends a system prompt, tools, MCP and memory files first. Measure it in Claude Code, Codex, Copilot CLI and OpenCode.

Dev.to | Oct 16, 2026 | Gabriel Anhaia

Automated excerpt

Claude Code defers MCP tool definitions by default: "Only tool names and server instructions load at session start" (MCP docs). Copilot CLI ships with the GitHub MCP server built in (GitHub docs). ENABLE_TOOL_SEARCH=false loads every MCP tool upfront (MCP docs).

Selected automatically from source text; not independently written or fact-checked. Read the original for full context.

Read the original article

Read next

AI briefing: recent picks

More stories to explore