#cost

Use /effort (low→max) or CLAUDE_CODE_EFFORT_LEVEL so simple edits stay cheap and hard debugging gets deeper reasoning without paying max on every prompt.

Inspect `/usage` (alias `/cost`) for per-skill, subagent, and MCP breakdowns, then cut the expensive outliers.

Every installed skill spends context just being listed, before you invoke anything. Sort the list by cost, then set the ones you never use to name-only.

Cycle into plan mode (Shift+Tab) so Claude explores and proposes an approach for approval before editing.

A skill that checks a link or reformats a file does not need your session model. Set model, effort and context: fork in the frontmatter so cheap skills stay cheap.

Shift specialized playbooks into on-demand skills so every session does not pay for instructions you rarely need.

Claude Code re-reads your whole conversation at the cached rate. Come back after a long lunch and the first message reprocesses all of it at full price.

Turn off MCP servers unrelated to the current project—their tool schemas still consume context at session start.

Add a top-level cache_control ephemeral flag so growing conversation prefixes reuse cached tokens and cut API cost and latency.