Switchyard lets LLM applications route traffic across models and providers while preserving native OpenAI and Anthropic API compatibility - enabling flexible model selection, benchmarking, and cost/performance optimization.
Rust
docs(routing): the classifier judge is not shown client instructions (#311)
ryan-lempka2a914728aa95
diff --git a/crates/libsy/src/algorithms/llm_class.rs b/crates/libsy/src/algorithms/llm_class.rsindex b9c86b62e..6367374dd 100644--- a/crates/libsy/src/algorithms/llm_class.rs+++ b/crates/libsy/src/algorithms/llm_class.rs@@ -71,8 +71,11 @@ impl TaskClassifierVerdict {}}-/// Keeps client instructions, the opening task, and the last `recent_turn_window`-/// turns after it. A window of `0` keeps the instructions and the task alone.+/// Keeps the opening task and the last `recent_turn_window` turns after it. A+/// window of `0` keeps the task alone.+///+/// Inbound decoders normalize client system and developer content into+/// `LlmRequest::instructions`, so it never reaches this list.////// Selects by reference and clones only what survives — a coding-agent/// conversation carries every tool result, so cloning it whole to keep a windowdiff --git a/docs/routing_algorithms/llm_classifier_routing.md b/docs/routing_algorithms/llm_classifier_routing.mdindex 1f87ab936..5bb000fcb 100644--- a/docs/routing_algorithms/llm_classifier_routing.md+++ b/docs/routing_algorithms/llm_classifier_routing.md@@ -100,7 +100,7 @@ for the server merge behavior.|---|---|---|| `base_threshold` | required | Lowest `p_solve` that routes a supported task to `weak_target`. Must be between `0` and `1`. || `threshold_step` | `0.0` | Amount added for each boundary step. Must be finite and non-negative, and `base_threshold + 2 * threshold_step` must not exceed `1`. |-| `recent_turn_window` | unset | When unset, the judge sees the opening user task and the latest user message when they differ. When set to `N`, it sees client system/developer instructions, the opening user task, and the last `N` conversation messages after that task. `0` keeps only the instructions and opening task. |+| `recent_turn_window` | unset | When unset, the judge sees the opening user task and the latest user message when they differ. When set to `N`, it sees the opening user task and the last `N` conversation messages after that task. `0` keeps only the opening task. Client system and developer instructions are not shown to the judge. || `session_affinity` | `false` | Retains the first selected target for a session and reuses it on later requests. || `message_hash_fallback` | `false` | When session metadata is absent, keys affinity from the first user-message text. Requires `session_affinity = true`. || `prompt` | packaged capability prompt | Replaces the classifier's system prompt. The packaged verdict schema and routing policy remain active. |