llm-usage-metrics 0.8.0 → 0.8.1

package/README.md

@@ -23,7 +23,7 @@
   <a href="./CONTRIBUTING.md">Contributing</a>
 </p>
 
-`llm-usage-metrics` reads local session data from 16 AI coding tools and converts it into one normalized usage history. Use it to review tokens and estimated cost, compare periods, find expensive sessions, correlate usage with local Git activity, or export the result.
+`llm-usage-metrics` reads local session data from 17 AI coding tools and converts it into one normalized usage history. Use it to review tokens and estimated cost, compare periods, find expensive sessions, correlate usage with local Git activity, or export the result.
 
 The CLI parses session content on your machine. It discovers standard source locations and includes a bundled pricing snapshot, so the first report can run without configuration or network access.
 
@@ -111,6 +111,7 @@ llm-usage events --since 2026-06-01 | jq '.totalTokens'
 | RooCode                | task JSON     |
 | KiloCode               | task JSON     |
 | Antigravity            | SQLite        |
+| DeepSeek Harness       | JSONL + zstd  |
 
 Each source adapter owns discovery and source-specific token normalization. Reports operate on the same `UsageEvent` shape after parsing. The [source documentation](https://ayagmar.github.io/llm-usage-metrics/sources/) lists default paths, override flags, and adapter-specific semantics.
 
@@ -185,7 +186,7 @@ Terminal, JSON, and Markdown availability varies by report. Usage, trends, wrapp
 
 ## Performance
 
-The repository publishes absolute cold and warm runtimes on real local corpora. The comparison includes the runs where `ccusage` is faster and the warm Claude run where the event ledger is faster. It also records the machine, commands, cache state, dataset size, and five-run distribution.
+The repository publishes direct-process and launcher-inclusive runtimes on stable snapshots of real local corpora. The current direct-process comparison shows `ccusage` ahead in every measured daily JSON cell; a separate monthly terminal check shows why timing `npx ccusage@...` can appear much slower than an installed `llm-usage` command. The benchmark records the machine, exact commands, application state, dataset size, and eight-run summary statistics.
 
 Read and reproduce the [benchmark](https://ayagmar.github.io/llm-usage-metrics/benchmarks/) before applying its results to another workload.