feat: Dispatch v2 Community Edition — full implementation#3
Open
nomadicmehul wants to merge 3 commits intomainfrom
Open
feat: Dispatch v2 Community Edition — full implementation#3nomadicmehul wants to merge 3 commits intomainfrom
nomadicmehul wants to merge 3 commits intomainfrom
Conversation
…ch v2 - Add ModelRouter with per-phase model selection (auto, provider-locked, pinned strategies) - Add model registry with cost tracking for Anthropic, Gemini, and GitHub Models - Add provider auto-detection from environment variables and CLI tools - Add MemoryManager with Tier 1 codebase context caching and Tier 2 cross-issue insights - Add ContextBudgetManager for token budget allocation across context sections - Implement real scoreConfidence() in all three engines (replacing stubs) - Add post-solve test verification that lowers confidence on test failures - Replace Promise.all with batched pipeline processing for cross-issue learning - Add --provider, --strategy, --no-memory CLI flags with backward compatibility - Add new config fields: provider, routingStrategy, enableCodebaseContext, enableCrossIssue
Smart model routing, memory system, cost tracking, OpenAI engine, checkpoint/resume, dispatch learn, and dispatch providers commands. - ModelRouter wired into pipeline for per-phase cost tracking - Cost breakdown by phase/provider in run summary - OpenAI engine (GPT-4.1, GPT-4.1 Mini, o3-mini) - dispatch providers and dispatch learn commands - Checkpoint/resume system (--resume flag) - Lessons from PR reviews injected into solve prompts - Version bump to v0.2.0-beta - CLAUDE.md and README.md updated for v2
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
Complete implementation of the Dispatch v2 Community Edition, delivering on the architecture proposal with smart model routing, a 3-tier memory system, cost tracking, and a polished CLI experience.
New Features
last-run.json--provider openaidispatch providers— diagnostic command showing detected providers, routing config, and registered modelsdispatch learn— scans Dispatch PRs for review feedback, extracts lessons to.dispatch/memory/lessons.jsonwith 30-day decaydispatch status --memory— shows memory state (context cache, lessons, checkpoint), provider config--resumeflag to continue from crashed runs; progress saved after each issuedispatch learninjected into solve prompts at low priorityv0.2.0-betaNew Files (7)
src/engine/openai.tssrc/commands/providers.tssrc/commands/learn.tstest/providers.test.ts,test/learn.test.ts,test/openai.test.ts,test/checkpoint.test.tsModified Files (14)
Stats
Test plan
npm run build— clean compilenpm test— 198 tests passdispatch providers— shows detected providers and routingdispatch status— shows last run with memory/cost infodispatch learn --show— shows lesson statedispatch run --dry-run— preview with routing infodispatch run --resume— resumes from checkpointdispatch run --provider openai— uses OpenAI engine