Cut my token bill 14 tools
Prompt Caching OptimizerPaste a prompt, get suggested cache_control spans + the boundary that maximizes hit rate.FREEModel RouterClassify the task, recommend the cheapest model that still nails it. Existing tool, lifted into the API bay.BYOKToken Counter + HeatmapCount tokens for any text on Haiku/Sonnet/Opus. Premium: per-paragraph heatmap + AI trim suggestions.BYOKImage Resizer + CompressorResize and compress images client-side to hit the dimension thresholds that cut token cost the most.FREEToken Receipt parserPaste an Anthropic usage block (any messages.create response). Get a human-readable cost breakdown.FREEConversation CompactorLong chat? Compact prior turns into a tight summary preamble before continuing. Premium: AI summary.BYOKPre-Send PreviewFinal eyes before send: token count, cost across models, cache-hit prediction. One last sanity check.FREECache-Aware Prompt BuilderBuild a prompt with explicit static / dynamic spans. See cache-write vs cache-read cost in real-time.FREEAuto-OptimizerPaste a prompt, get a rewritten version with a verifiable % savings + side-by-side diff + semantic-match score.Operator+Output Length LinterLint your max_tokens settings vs measured output lengths. Cut over-provisioned ceilings.Commander+Project Instruction CompressorCompress bloated project instructions without losing intent. Premium: Claude validates the rewrite.Operator+Off-Peak SchedulerSurface batchable Claude tasks. Suggests a daily off-peak window when output is fast and rate-limit risk is low.Operator+/clear /compact CoachRecommend the right /clear vs /compact moment based on session length, drift, and token spend.Commander+Session Cost TrackerLive per-session cost. Wired to /mc-cost via the CLI hook. Real-time, not after-the-fact.Commander+Sign in to pin tools, save results, and build your own workbench.