Add wave-based parallel subtask execution to /map-efficient#80
Merged
Conversation
Remove mem0 MCP integration entirely from the project: - Delete Curator agent (curator.md) and all mem0 tool references - Clean all 8 agent templates, 7 command files, docs, skills, tests - Remove ACE learning system references from plugin metadata - Delete obsolete files (tier3_integration_report.json, apply_tier3_deltas.py, playbook.json.backup, requirements-semantic.txt) - Update __init__.py to remove curator generator and mem0 allowlist Simplify map-review to always use thorough mode: - Remove BIG/SMALL mode selection menu - Always present top 4 issues per section Fix hook errors causing noise during agent work: - block-secrets.py: add SAFE_PATH_PREFIXES to stop blocking .claude/ framework files - safety-guardrails.py: narrow "token"/"password" patterns to file extensions only, add .claude/ dirs to safe prefixes, fix missing execute permission - settings.local.json: remove duplicate workflow-context-injector.py (matcher "*", timeout 1s) and duplicate end-of-turn.sh Stop hook All templates synced via make sync-templates. 635 passed, 1 skipped, 0 failures.
…ution to /map-efficient Introduce a "wave" concept for executing independent subtasks in parallel. The task-decomposer already outputs dependency DAGs; this change computes topological waves (Kahn's algorithm) and splits waves by file conflicts, enabling Actor and Monitor agents to run concurrently for independent subtasks. Changes: - dependency_graph.py: Add compute_waves() and split_wave_by_file_conflicts() - map_orchestrator.py: Add set_waves, get_wave_step, validate_wave_step, advance_wave commands with StepState wave fields (backward compatible) - map_step_runner.py: Add update_workflow_state_batch() for multi-subtask updates - workflow-gate.py: Support active_subtasks list for parallel wave mode - workflow-context-injector.py: Show wave progress in status line - map-efficient.md: Document wave computation and parallel execution loop - Tests: 50 new tests covering waves, file conflicts, orchestrator commands, and backward compatibility (668 passed, 1 skipped total) Linear DAGs degrade naturally to single-subtask waves (identical to current behavior).
|
Caution Review failedThe pull request is closed. ℹ️ Recent review infoConfiguration used: defaults Review profile: CHILL Plan: Pro 📒 Files selected for processing (93)
📝 WalkthroughWalkthroughThis PR removes the Curator agent and eliminates the ACE Learning System with mem0 MCP (memory pattern storage), reducing the agent count from 12 to 11. All mem0-based pattern searches, storage operations, and Curator-mediated learning are removed across documentation, templates, and configs. Concurrently, wave-based parallel task execution and dependency graph analysis are introduced to the orchestration layer. Changes
Estimated code review effort🎯 4 (Complex) | ⏱️ ~60 minutes Possibly related PRs
Poem
✨ Finishing Touches
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
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
/map-efficientworkflowcompute_waves()(Kahn's algorithm) andsplit_wave_by_file_conflicts()(greedy coloring) toDependencyGraphset_waves,get_wave_step,validate_wave_step,advance_wavewith backward-compatibleStepStatefieldsupdate_workflow_state_batch()for multi-subtask state updates in a single callworkflow-gate.pyto supportactive_subtaskslist for parallel wave modeworkflow-context-injector.pyto display wave progress in status linemap-efficient.mdHow it works
The task-decomposer already outputs
dependencies: [...]per subtask forming a DAG. This PR computes topological waves — groups of subtasks whose dependencies are all satisfied by prior waves. Within a wave, Actor and Monitor agents run in parallel via multipleTask()calls in a single message.Example:
ST-001(no deps) →ST-002(deps: ST-001) +ST-004(deps: ST-001) →ST-003(deps: ST-002, ST-004)Produces waves:
[["ST-001"], ["ST-002", "ST-004"], ["ST-003"]]Linear DAGs degrade to single-subtask waves (identical to current sequential behavior).
Test plan
TestComputeWaves: linear chain, fan-out, diamond, cycle detection, empty graph, multiple roots, dangling deps (8 tests)TestSplitWaveByFileConflicts: no overlap, partial overlap, all overlap, empty files, missing from map (7 tests)TestSetWaves: correct waves from blueprint, state persistence, missing blueprint, file conflict splitting (4 tests)TestGetWaveStep: parallel/sequential mode detection, completion, no-waves fallback (4 tests)TestValidateWaveStep: phase advancement, evidence gating (3 tests)TestAdvanceWave: index increment, completion, phase reset, error handling (4 tests)TestBackwardCompat: sequential flow without waves, state serialization, old state file loading (3 tests)Summary by CodeRabbit
Release Notes
New Features
Documentation
Removals