Most Used Tags
Capture ephemeral notes during multi-agent workflows for better context sharing.
Architect evaluates system architecture and module integrity without modifying code.
Quickly execute trivial tasks inline without overhead or planning.
Code review specialist that inspects code quality before approval.
Systematic debugging agent that follows a strict investigation protocol.
Maintains the quality and relevance of the learning playbook by integrating and pruning lessons.
Verify code implementations against original task goals to ensure accuracy.
Lesson extraction specialist that captures actionable insights from completed workflows.
devt is a multi-agent development workflow plugin for Claude Code, streamlining coding, testing, and documentation.
Interpret and triage architecture scanner output to produce a prioritized remediation plan.
Scan the codebase for existing implementations to prevent duplication before writing new code.
Evaluate code quality with a structured 100-point scoring framework and actionable findings.
Assess task complexity before starting work to ensure proper planning and execution.
Quickly search the learning playbook for past lessons before implementation to avoid repeating mistakes.
Verify that completed implementations are genuine and connected, not just stubs or dead code.
Generate weekly development reports from git history, summarizing team activity and contributions.
Capture actionable lessons from completed work to enhance future performance.
Easily check for and install updates for the devt plugin from GitHub.
Interactive project setup wizard for configuring development environments.
Enforce Test-Driven Development by writing failing tests before production code.
Create a validated implementation plan before coding by analyzing tasks and breaking them into steps.
Automate the complete development workflow by detecting task complexity and executing the appropriate pipeline.
Detects recurring correction patterns to propose improvements for the devt plugin system.
Manage persistent context threads for cross-session work and investigations.
Testing specialist for writing, updating, and validating tests in workflows.
Diagnose the health of your devt plugin by checking configuration and state integrity.
Maintain the quality of your learning playbook by deduplicating, resolving contradictions, and improving entries.
Run quality gates to ensure code integrity through linting, type checking, and testing.
Capture ideas instantly without interrupting your workflow. Promote notes to tasks later.
Displays current workflow status, including completed steps and next actions.
Route freeform text to the appropriate development command effortlessly.
Clarify implementation choices and capture decisions before coding to avoid assumptions.
Pause ongoing workflows and create structured handoff files for seamless session resumption.
Implementation specialist for writing and modifying code in workflows.
Abort active workflows and reset their state to idle.
Perform architecture health scans to identify violations and structural issues in your codebase.
Create detailed session reports summarizing work done, decisions made, and outcomes achieved.
Comprehensive guide to all devt commands, from basics to advanced use cases.
Systematic debugging using a 4-phase investigation process to identify and fix issues.
Diagnose failed or stuck workflows through post-mortem analysis of artifacts and git history.
Automatically create pull requests with detailed descriptions from workflow artifacts.
Automate the creation and updating of module documentation after code changes.
Technical investigation specialist for planning unfamiliar features.
Streamline your coding process by skipping documentation and retrospectives for quick implementations.
Fetch the latest documentation for any external library before coding to avoid outdated information.
Generate detailed Product Requirements Documents (PRDs) through systematic analysis and interviews.
Automatically detects your current workflow state and executes the next logical step.
Conduct a thorough, read-only code review with actionable insights and recommendations.
Investigate implementation approaches before planning to avoid pitfalls and identify patterns.
Facilitates informed decision-making by comparing multiple approaches with trade-off analysis.
Archive stale lessons from the learning playbook to maintain high-value knowledge.
Extract valuable lessons learned from current sessions into persistent memory.