Source Intelligence
Reading a new release v2.1.251 Analysing changes · 2/5 Deeper second pass · 0/5 agents 427 findings $12.75 so far

DisclaimerUnofficial, and not affiliated with Anthropic. Nearly all of this is read straight out of what ships: npm bundles, captured prompts, published docs. Anthropic's own notes go in verbatim, marked as theirs. The rest is my reading, and every entry carries the strings behind it. If one looks wrong, vote it down and say why.

All of v2.0.20 Home All releases olderv2.0.19 v2.0.21newer
Claude Code v2.0.20

Context Usage Markdown Export

What: Export detailed context usage statistics as markdown tables for non-interactive sessions

How to use:

# Run context command in non-interactive mode
claude --output text context

# Output includes markdown tables:
# - Categories (with tokens and percentages)
# - MCP Tools
# - Custom Agents
# - Memory Files
# - SlashCommand Tool statistics

Details:

  • Token counts formatted for readability (e.g., "1.5k" instead of "1500")
  • Includes model information and total usage percentage
  • Shows free space and autocompact buffer separately
  • Agent sources identified (Built-in, Project, User, Local, Flag, Policy, Plugin)
  • Evidence: Ma6() at line 301494, rk() token formatter at line 301491

See this entry in the whole of v2.0.20 →