korchasa/flow-answer

Task Answer

Overview

Analyze and answer a user's question in Autonomous Mode by carefully reading documentation and inspecting the codebase. For large responses, save detailed analysis to documents/whiteboard.md.

Context

<context> The user has a question about the project's logic, architecture, or implementation. The project uses a structured documentation system in `./documents` and follows specific design patterns. </context>

Rules & Constraints

<rules> 1. **Language Policy**: - Code, technical terms, and quotes from codebase: **English**. - Analysis, explanations, and chat responses: **User's Query Language** (or Russian if undefined). 2. **Read-Only Mode**: - Keep all repository files unchanged (except `documents/whiteboard.md`). - Use tools in read-only mode (reading files, searching code). - **DO NOT** modify project configuration or code. 3. **Accuracy**: Cross-check implementation against requirements and design to identify matches, gaps, or inconsistencies. 4. **Planning**: The agent MUST use a task management tool (e.g., todo write) to track the execution steps. </rules>

Instructions

<step_by_step>

1. Initialize
   • Use a task management tool (e.g., todo write) to create a plan based on these steps.
2. Understand the question
   • Read all available docs in ./documents (SRS, SDS, file structure, manuals, whiteboard) if they exist.
   • Clarify and restate the user's question in a more precise and structured form.
   • Ask the user follow-up questions if the task, constraints, or expected output are not fully clear.
   • Explicitly list assumptions, missing details, and the intended scope of the answer.
3. Documentation-based analysis
   • Map the question to relevant requirements in documents/requirements.md (SRS).
   • Map the question to relevant design decisions in documents/design.md (SDS).
   • Consider architecture, constraints, and interfaces described in the documentation.
4. Code analysis
   • Locate relevant files, modules, and functions in the codebase.
   • Read and understand current implementation and its behavior in the context of the question.
   • Identify matches, gaps, or inconsistencies between implementation and documentation.
5. Answer synthesis
   • Explain how the existing code and design relate to the question and requirements.
   • Highlight limitations, edge cases, and potential improvements where relevant.
   • If critical uncertainties remain, clearly state them and ask for input instead of guessing.
   • For large responses (>1000 characters), save detailed analysis to documents/whiteboard.md and provide summary in chat. </step_by_step>

Verification

<verification> - [ ] Documentation read (where available) and question clarified. - [ ] Necessary clarification questions asked to the user (if needed). - [ ] Relevant code located and analyzed. - [ ] Answer formulated and structured. - [ ] Response saved to `documents/whiteboard.md` if large, otherwise provided directly. </verification>