Stabilize v8-engine rewrite: engine hardening, index health, desktop fixes - #39
Merged
Merged
Conversation
…ce directory validation
…iagnostics - Added support for tree-sitter syntax checks via a new VerificationSyntaxPort. - Implemented logic to prefer syntax checks from the port over traditional methods. - Enhanced the verification pipeline to utilize script hints from AGENTS.md for better check prioritization. - Introduced new diagnostics handling for syntax errors, allowing for more granular feedback. - Added tests to ensure proper functionality of syntax checks and their integration into the verification process.
- Implemented the git_signoff_range tool to add Signed-off-by trailers across a commit range using git rebase. - Introduced input and output schemas for the tool. - Added necessary validations for tool grants and protected branches. - Enhanced the execution flow to handle stashing and restoring changes during the rebase process. - Updated decision policy to recognize vcs_history_rewrite requests and prefer git_signoff_range over apply_patch for commit metadata changes. - Added tests for detecting DCO-related prompts and ensuring proper handling of git push arguments. - Updated documentation and constants to include the new tool and its implications on decision-making processes.
…ing and attachment normalization - Introduced new schemas and types for handling meta commands, turn kinds, and session actions. - Added functionality to classify leading slash commands and extract mention artifacts from user messages. - Implemented attachment normalization to enforce mime type and size constraints. - Enhanced the RequestIntakePipeline to process sanitized messages, classify commands, and build user request envelopes with additional fields. - Updated README documentation to reflect changes in module functionality and structure. - Added tests for new features including command classification, mention extraction, and message sanitization.
…ce request intake processing
- Enhanced RuleIntentClassifier to handle multiple task matches with a weak heuristic for SuperIntent. - Introduced new patterns for bugfix intent in RulePatterns to capture more specific requests. - Updated isWholeRequestReadOnlyConstraint to include new mutation verb checks. - Improved TurnKindIntentPolicy to promote plan approvals to act based on user messages. - Added functionality in SuperIntent to prioritize strong rule classifications over weaker LLM votes. - Refined TaskAnalyzer to better handle continuation turns and clarify intent confidence. - Updated tests to cover new behaviors in intent classification and task analysis.
- Added discrete grant profiles to manage tool permissions based on mode and route. - Implemented `selectGrantProfile` function to determine the appropriate grant profile. - Enhanced `buildToolGrant` to utilize the new grant profiles and streamline tool permissions. - Refactored related functions for clarity and maintainability. - Updated decision policy to incorporate new logic for handling network authority and tool grants. - Added unit tests for grant profiles and decision pipeline integrity.
- Introduced `injectSkillCatalogL1` flag to enable injection of a name+description awareness strip for skills. - Added `skillCatalogL1` parameter to pass entries for the L1 catalog. - Updated relevant types and schemas to accommodate new catalog structure. - Enhanced `SkillsPipeline` and `PromptConstructionPipeline` to handle catalog generation and injection. - Implemented `SkillCatalogFragment` for rendering the L1 catalog in prompts. - Added tests to verify the functionality of the L1 skill catalog injection.
…lance and syntax checks
- Added `listPriors` method to `RuleIntentClassifier` for heuristic and explicit intent classification. - Introduced `UnderstandingEvidencePack` schema and related types for structured evidence collection. - Implemented `buildUnderstandingEvidencePack` function to create evidence packs from user input and classifier results. - Developed utility functions for computing task size and analyzing message characteristics (e.g., paste dumps, test failures). - Updated `RequestUnderstandingPipeline` to integrate evidence packs and task size into the understanding process. - Enhanced `RulewiseTaskAnalyzer` to utilize task size and planning hints for improved task analysis. - Added tests for evidence pack generation and task size computation. - Refactored related schemas and types to support new evidence structures and planning hints.
- Introduced a fallback message for incomplete assistant turns to improve user experience. - Added functions to evaluate the usability of prior answers and synthesize fallback responses based on changed files. - Implemented a new module for mutate readiness, including budgeting and readiness evaluation for active checklist steps. - Enhanced mutation nudge logic to incorporate plan drafting and escalate readonly thrash situations. - Updated the model loop to integrate new readiness checks and handle incomplete answer recoveries more effectively. - Added tests for the new mutate readiness functionalities and mutation nudge behaviors. - Refactored existing code to improve clarity and maintainability, including the removal of deprecated methods.
… readiness management
…nced LLM management
Relocate the Mitii brand mark from the activity bar and history panel into the app top bar, and surface a New chat action in a dedicated chat top bar when code mode is active.
- Introduced `readIndexPipelineHealth` function to assess the health of indexing pipelines (Code, FTS, Embeddings, Graph, Map, Tree-sitter, and native modules). - Enhanced `getIndexStatus` to include pipeline health information in the status response. - Updated `IndexStatusChip` component to display detailed pipeline health status. - Modified `App` and `SettingsPanel` components to utilize the new health data. - Added CSS styles for displaying pipeline health information. - Implemented tests for the new health tracking functionality. - Updated documentation to reflect changes in index pipeline health reporting.
…urfacing Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
…xtWindow Co-authored-by: Cursor <cursoragent@cursor.com>
codewithshinde
changed the title
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
developmentElectron launcher (run-electron.cjs) while keeping the 2.10.x version line.ARCHITECTURE.mdorchestration ownership now points atv8-engine.enabled;contextWindowtests import vitest correctly.Test plan
pnpm --filter @mitii/v8 run test:v8-engine(211 passed)pnpm --filter @mitii/sdk build && pnpm --filter @mitii/sdk test(17 passed)pnpm --filter @mitii/desktop typecheck && pnpm --filter @mitii/desktop test(85 passed)pnpm --filter @mitii/desktop start→ open workspace → reindex → confirm logs under storage logs path