token-accounting

Independent verification of AI support bills: applicable terms, invoice and business records, with reviewable findings. Open-source conformance for agent telemetry — PASS / FAIL / UNPROVABLE in CI.

Bounded GenAI metrics and keyed summaries from OpenTelemetry traces

A prefix-cache profit-and-loss layer for DeepSeek coding agents — wrap your client in two lines, see per-request cache HIT/PARTIAL/MISS, the ¥ each prefix-bust wasted, and the reorder that restores the discount.

Deterministic benchmark for agent instruction adherence, with token and round-trip cost verified by a recording proxy

Local, deterministic, report-only Token ROI and multi-agent worktree accounting

Embeddable LLM benchmark speed measurement and offline Codex/Claude Desktop analysis, with explicit timing boundaries and output verification.

Trace creator asset delivery token use and processing exceptions with one Java configuration.

Conformance invariants and runnable harnesses for tools that count LLM tokens and cost from agent transcripts. Two catalogs, eleven upstream fixes.