v0.121.4

## What's Changed
* fix(frontend): stop a palette scroll race from failing web unit tests by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7184
* fix(observability): correct tracing analytics figures by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7176
* fix(sdk): keep one context across streamed pulls and pin usage to the workflow span by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7175
* fix(api): roll up costs and tokens for agent batches, zero costs and summary spans by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7179
* fix(api): recompute trace totals when a trace arrives in several OTLP requests by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7181
* fix(api): price LLM spans correctly and mark spans that cannot be priced by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7178
* fix(runner): correct token and cost reporting for the Pi, Claude and Codex harnesses by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7177
* fix(runner): keep usage across an approval pause, and mark custom-connection model spans by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7183
* fix(runner): mark spans served by a custom model connection by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7187
* chore(runner): bump the Pi harness to 0.87.1 so Pi prices calls with current rates by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7180
* feat(runner): use OpenRouter's billed cost on Pi chat spans by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7185
* fix(mobile): fill the trace drawer's data slots so tabs show span data by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7182
* fix(frontend): wait for the build kit before the first /m message by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7190
* fix(api): keep chat app inputs when an evaluation invokes it by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7196
* fix(api): download litellm's latest price list at startup again by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7197
* fix(runner): keep a late Claude cost reading from overwriting the turn's usage by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7203
* [release] v0.121.4 by @github-actions[bot] in https://github.com/Agenta-AI/agenta/pull/7191


**Full Changelog**: https://github.com/Agenta-AI/agenta/compare/v0.121.3...v0.121.4