Tools for teams running agents in production
All free, no signup, and no data leaves your browser — they are static pages with JavaScript, so there is nowhere for it to go. Built because I needed them for client work and it was easier to make them public than to keep rebuilding them.
Assessment· 4 minutes
Agent Production Readiness Scorecard
Twelve questions that separate an AI agent demo from a system you can bill customers for. Scored across reliability, cost control, and observability.
Open it- 12 weighted questions
- Scored 0–100 across three pillars
- Gaps ranked by production risk
- Runs entirely in your browser
Calculator· 2 minutes
Agent Cost Calculator
Model monthly LLM spend from volume, tokens, retries and cache hits — including the tool-definition tokens and tokenizer differences most calculators ignore.
Open it- Provider pricing verified 2026-08-02
- 8 models compared side by side
- 5 realistic workload presets
- Ranked savings scenarios with effort estimates
Reference· Browse
AI Agent Failure Taxonomy
Nine ways production AI agents fail, each with the root causes worth checking first and the fixes that actually hold. Named categories, because failures you cannot name are failures you cannot count.
Open it- 9 named failure modes
- Root causes with a diagnostic check for each
- Fixes with effort estimates
- Filterable by pillar and severity
Something missing?
If there is a calculation or reference you keep rebuilding by hand, tell me — I will probably build it, because I likely need it too.