execute-prp

Prompt file imported from abuklea/prompthub (.claude/commands/execute-prp.md). Fill in {{arguments}} before use. Copyright stays with the author.

Execute BASE PRP

Implement a feature using using the PRP file.

PRP File: {{arguments}}

Execution Process

1. Load PRP

   • Read the specified PRP file
   • Understand all context and requirements
   • Follow all instructions in the PRP and extend the research if needed
   • Ensure you have all needed context to implement the PRP fully
   • Do more web searches and codebase exploration as needed

2. ULTRATHINK

   • Think hard before you execute the plan. Create a comprehensive plan addressing all requirements.
   • Break down complex tasks into smaller, manageable steps using your todos tools.
   • Use the TodoWrite tool to create and track your implementation plan.
   • Identify implementation patterns from existing code to follow.

3. Execute the plan

   • Execute the PRP
   • Implement all the code

4. Validate

   • Run each validation command
   • Fix any failures
   • Re-run until all pass

5. Complete

   • Ensure all checklist items done
   • Run final validation suite
   • Report completion status
   • Read the PRP again to ensure you have implemented everything

6. Reference the PRP

   • You can always reference the PRP again if needed

Note: If validation fails, use error patterns in PRP to fix and retry.

Imports

@docs/rules/project.md @docs/rules/archon.md @docs/rules/documentation.md @docs/rules/documentation-prp.md

Important Notes

• FIRST STEP: Ensure the PRP start report has been created and is accurate and current. If it does not exist, create it
• CRITICAL: Follow the definitive Archon process
• Task status flow: todo → doing → review → done
• Keep queries SHORT (2-5 keywords) for better search results
• Higher task_order = higher priority (0-100)
• Tasks should be 30 min - 4 hours of work
• Split tasks or related collections of tasks, and factor into smallest discrete self-contained tasks (where possible and optimal)
• Parallelize tasks where possible using subagents and the Task tool
• ALWAYS use the most aligned/suitable subagent available for any discrete task or collection of related tasks.
• Keep the Archon tasks list up to date, and update task statuses - set tasks to "doing" before beginning implementation
• Move Archon tasks to review before completion
• Make sure that Archon tasks are synched with the PRP documents, and all Archon titles are prefixed with the correct Phase and Task IDs PXTX (e.g. P1T1: Initialise project)
• Use subagents or subagent teams, to assign tasks to the best/optimal, most-suitable, and most-qualified agent/s.
• Spawn agents in parallel using the Task tool, and when tasks have been marked as parallelizable [P]
• Share comprehensive and detailed information between agent/caller to guarantee robust results and context integrity and relevance.
• NEVER create duplicate tasks - ALWAYS check existing tasks first to ensure new tasks are unique and not duplicating task (or subtask) entries. Update existing tasks where suitable.
• For each Archon task, identify discrete tasks or collections of related tasks (and sequences of tasks) that are suitable for parallelization, and mark each task with [P] (this will be used to schedule and run multiple subagents simultaneously using the Task tool).
• Make sure the assignee of each task is set to the most-suitable/most-aligned agent to complete each task.
• Provide a lean but complete foundation for the task within the task description:
• A clear task goal and success criteria, with error handling and fallback guidance
• Extensive relevant details and context to complete the task
• Indicate any task dependencies or pre-requisite resources, configurations, or prior-work
• Suggest if suitable for parallelization (mark task title with [P]) , and any particular task sequence requirements
• Specify intended task sequence, scope, and code-base influence
• Include references to any required/useful project documents, or external sources
• Instruct agents to share markdown files if necessary, written to folder wip/ - clearly named, also with agent name and datetime

CORE INSTRUCTIONS

  ✅ Use the Task tool to spawn multiple subagents to work in parallel.   ✅ NO mock/dummy data - Only real, validated data.   ✅ NO placeholder components, systems, code, or data to be used EVER! - Real and actual data/implementations only.   ✅ NO backwards compatibility - Clean, modern implementation.   ✅ Real database operations ALWAYS - No bypassed or disabled features.   ✅ Production-ready code ONLY - No temporary workarounds EVER!   ✅ NEVER ASSUME, PRESUME, OR GUESS - When in doubt, ask for clarification.   ✅ ALWAYS require a full and complete implementation AND integration of every feature into the platform - All subtasks, and task implementations/solutions/outputs MUST be fully complete, robust, and integrated and tested successfully, AND the feature/solution working correctly broadly within the platfom, BEFORE a task is to be considered completed or the task status is set to review.   ✅ NEVER assume task completeness - or over-estimate progress or completion status until the developer has reviewed and tested the implementation.     ✅ Always VERIFY file paths and module names before use.   ✅ Keep CLAUDE.md updated when adding new patterns or dependencies.   ✅ TEST your code - No feature is complete without passing tests.   ✅ DOCUMENT your decisions - for the dev team, and for future developers and contributors.   ✅ NEVER "WORK AROUND" anything! - Real, correct, and targeted solutions only are permitted!   ✅ REMINDER: NEVER USE OR ADD PLACEHOLDER COMPONENTS/SYSTEMS/CODE/DATA