ClaudeHack

Context Management

9 hacks

  1. 1.

    Ask Claude to log architecture decisions and TODOs into a repo file that can be re-read after /compact.

    0.00 points·3 months ago·easy·discuss
  2. 2.

    Proactively run `/compact` with focus instructions so summaries keep the work that matters instead of stale debugging.

    0.00 points·3 months ago·easy·discuss
  3. 3.

    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.

    0.00 points·3 months ago·easy·discuss
  4. 4.

    Double-Esc /rewind back to after useful file reads, then re-prompt with what you learned—without the failed approach still in context.

    0.00 points·3 months ago·easy·discuss
  5. 5.

    Run /context to break down token consumers—CLAUDE.md, MCP tools, conversation—and prune the expensive ones.

    0.00 points·3 months ago·easy·discuss
  6. 6.

    Declare a Compact Instructions section in CLAUDE.md so both manual /compact and autocompact preserve the project’s critical decisions by default.

    0.00 points·3 months ago·easy·discuss
  7. 7.

    Use /clear between unrelated tasks so leftover debugging does not bias the next feature.

    0.00 points·3 months ago·easy·discuss
  8. 8.

    Use /btw for single-turn clarifications that share full context but never enter conversation history or burn future tokens.

    0.00 points·3 months ago·easy·discuss
  9. 9.

    Auto-compact waits until your context is nearly full, which is well past the point where answer quality starts sliding. Set the window lower and compact while the session is still sharp.

    0.00 points·3 months ago·easy·discuss