mracal

Capture and log architectural decisions automatically from code changes and conversations.

Plan and document rollback procedures for risky changes to ensure smooth recovery.

Generates adversarial and spec-driven test cases to ensure code robustness.

Organize and document resolved bugs to prevent future issues.

Validates requirements against SMART criteria to catch vagueness early.

Dispatch research agents to gather insights before proposing features or architectures.

Enforces structured output for analyses, reports, and recommendations to enhance clarity and comprehension.

Bidirectional sync between specification documents and code to prevent spec drift.

Enforces a rigorous quality loop to ensure clean code before declaring work done.

Generate a behavioral scenario acceptance matrix for comprehensive test coverage planning.

Codify recurring code patterns for efficient reuse in your projects.

Generate session continuity documents to maintain context across long development sessions.

Prevents shortcut rationalization in LLMs by detecting when steps are skipped or requirements are simplified.

Monitor the health status of all configured services, including uptime and errors.

Conducts comprehensive gap analysis to assess production readiness of codebases against specifications.

Conducts thorough end-of-iteration reviews with quality grading and trend analysis.

Automates the closure of production readiness gaps through iterative analysis and execution.

Challenge recommendations and designs through adversarial review to identify weaknesses before production.

Sets quality expectations for production implementation, ensuring robustness and reliability.

Scan staged changes for common commit mistakes before finalizing a git commit.

Validates completed work against defined acceptance criteria to ensure tasks meet success standards.

Generates fuzz test scaffolding for parsers to identify unexpected input crashes.

Directs users to the appropriate vibe-engineering skill based on their current task.

Decomposes large tasks into independent subtasks for parallel execution, optimizing efficiency.

Robustly implement snapshot/golden file tests with temporal normalization to prevent daily breakage.

Systematically addresses large backlogs of issues using prioritized waves.

Enforces tiered test coverage standards across line coverage, spec-to-test, and spec-to-code dimensions.

Extracts learnings from completed work and updates a persistent learnings file.

Identify discrepancies between specification documents and implementation code to ensure compliance.

Prevent scope creep during implementation by catching unrequested features and unnecessary changes.

Safely deploy applications with pre-flight checks and automatic rollback.

Safely integrates commits from parallel agent branches using sequential cherry-pick.

Manages context window limits by summarizing and compacting session state for better coherence.

Manage non-blocking tasks with a persistent cross-session queue.

Conduct thorough research before making design decisions to avoid common pitfalls.

vibe-engineering provides 37 practical skills to enhance engineering workflows in Claude Code.

Audits tests for concurrency safety to prevent race conditions and ensure reliable test outcomes.

Bootstraps the agent to utilize vibe-engineering skills for enhanced development practices.

Quickly checks technical documents for structural quality with a 6-point assessment.