03ba79d261
* feat(lens): add lens_feedback ClickHouse table for human trace scores ReplacingMergeTree(UpdatedAt, IsDeleted) keyed like agent_traces_by_key so feedback joins traces in sort order, with a bloom filter on TraceId, a score CHECK, and the same retention as traces Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(lens): add scoped ClickHouse reads for trace feedback feedback_target resolves a visible trace's team, key and trace_ref before a write; feedback lists the latest live entry per author; feedback_summary returns count, average and lowest score for a batch of traces Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * test(lens): pin the Claude session to trace id hash shared with feedback Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(lens): expose feedback queries to the Python trace bridge Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(lens): add typed feedback models and ClickHouse feedback store Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(lens): add /lens/feedback API for 0-10 human scores on traces PUT saves or replaces the caller's score and comment, GET lists every entry, DELETE removes the caller's, POST /summary feeds the trace list. Callers can target a trace_id or a Claude Code session_id Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(lens-ui): add typed feedback calls to the traces API Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(lens-ui): flag rated traces and filter the list by feedback A Feedback column shows each run's average score and rating count, marked red when any score is 4 or below, and a filter narrows to rated or low-score runs through the URL Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(lens-ui): show and edit human feedback inside a trace Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(lens): let app keys post end-user feedback on their own traces Any key can now PUT/DELETE feedback on traces its team or key sent, naming the end user in an optional user field. Reading feedback stays with Lens admins Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(lens-ui): show end-user feedback first in a run and flag low-scored rows Replace the edit popover and feedback filter with a read-only view: the list shows each run's score in a fixed-width badge and tints the row red when a user scored it 4 or below, and opening a run shows what users said with their score right under the header Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> |
||
|---|---|---|
| .. | ||
| schemas | ||
| templates | ||
| config.json | ||
| README.md | ||
Run uv run scripts/generate_trace_types.py from the repository root to export Rust schemas and regenerate the Python trace contracts. Run uv run scripts/generate_trace_types.py --check to compare fresh output with the committed schemas and Python files
The script pins datamodel-code-generator in its inline dependency metadata. Rust uses the workspace's locked Schemars version through each owning crate's optional schema feature. Neither tool is a Python runtime dependency
The litellm-traces Rust request types own the generated request models in litellm/rust_bridge/trace/generated/requests.py. The GET routes bind their query parameters directly to the generated models. GET request types allow unknown fields because existing clients' unknown query parameters are ignored. The SQL body model forbids extra fields
Each crate exports its own roots using JSON Schema 2020-12. Request parameters use Schemars' deserialization contract and carry only explicitly declared constraints. Their schemas skip the integer-bounds transform because it would add i64 bounds to start_ms and end_ms, narrowing what Python accepts, and replace page_size's explicit 1..500 range with 0..65535. Trace views and query help use the serialization contract. Lens rows use their ClickHouse deserialization schemas, including quoted numbers and numeric boolean flags
The templates preserve tuple conversion, immutable tuple defaults, and bounded ReadOnly TypedDict fields. Pydantic models use the generator's frozen-model option and each schema's extra-field policy. ClickHouse numeric schemas select bounded, normalized Python scalar types through schema metadata consumed by the model template
Changing the public request contract requires a separate behavior-change PR
Edit the owning Rust contract, schema annotation, or generation configuration, then regenerate. Never edit litellm/rust_bridge/trace/generated/ manually. responses.py holds the generated SQL response, and queries.py keeps the ClickHouse envelope as an internal validator