Classification: LIVE tool capabilities are inspectable; usage, adoption and outcome impact are UNKNOWN.
View the full infographic → · View this article
The tools were built around recurring questions in Token Optimisation’s practical workflow work: what an agent should do when state or access changes, which context rules are selected, and how a cost scenario should expose assumptions.
Three narrow jobs, three separate pages
- AI Agent Operating Rules Builder: select safeguards for checkpoints, outages, idempotency, human gates, crash recovery and progress reporting.
- AI Token Optimisation Rules Builder: choose context, routing, reuse and measurement rules and see the resulting Markdown.
- AI Context Cost Calculator: compare official API price presets or enter subscription measurements; missing denominators stay “not calculated.”
What they do not prove
Interactive output can be checked in the browser, but the existence of a tool does not prove token savings, better jobs, adoption or reduced review time. Those claims require usage evidence and a comparable acceptance test.
Each page explains its inputs and limitations. Calculations run locally and do not call an AI, pricing or FX service. API rates carry source links and dates; the Pro 100 charge is user-recorded, not a subscription token rate.
Start from the tools directory or the scoped observations in request measurement and recovery signals.