Update ADR-0003 to reference continuous improvement schema#162
Merged
Conversation
Adds a 'Related' section documenting how the CI schema (evaluation-driven refinement) relates to tiered cognition: per-attempt trace data feeds cognition_summary, refinement amortizes strong-model planning across cheap-model attempts, and no CognitionTier enum is introduced (consistent with ADR-0003 D1 / ADR-0002 G1). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Adds a Related section to ADR-0003 documenting the relationship between tiered cognition and the continuous improvement schema.
Three connection points:
Trace data —
LineageAttempt.replay_metadataincludesruntime_binding_model, feeding directly intocognition_summary.nodes_by_model(ADR-0003 D2). A CI run withmax_attempts=3gives paired data points with consistent evaluation criteria — exactly the evidence G4 requires before a routing rule is safe.Bounded-cognition amortization — CI's
RefinementPolicymirrors the "plan once, execute many" pattern from ADR-0003 Context. Strong model sets theEvaluationSpecbaseline once; cheap/local model executes attempts; scoring loop provides feedback.No
CognitionTierenum — explicitly noted, consistent with ADR-0003 D1 and ADR-0002 G1.Test plan
🤖 Generated with Claude Code