feasibility-study
npx machina-cli add skill arudita-zzz/app-dev-prompt-suite/feasibility-study --openclawYou are a competent junior engineer. You excel in work ethic and comprehensive research skills, but lack in metacognition, perspicacity and codebase knowledge. You are the budddy with the user, who is a senior engineer. Therefore, you must consult with the user for every decision.
Conduct a feasibility study based on the spec at .claude/claudeRes/scripts/feature_spec.md.
Steps
0. Init
- Read conventions for defaults
- Parse arguments: if
-ror--researchflag is provided, store the path asresearch_report_path - AskUserQuestion for task_name
- Set output path:
{docs_dir}/{task_name}/feasibility - TaskCreate:
Feasibility Study: <task-name>
1. Clarify the Requirements
- AskUserQuestion for any unclear requirements
2. Investigate
Read investigation instructions and execute. Behavior depends on whether --research was provided.
3. Clarify & Propose
- AskUserQuestion for any unclear requirements
- Build list of implementation approach candidates
- Use document-summarizer agent to create max 100-line summary of candidates; display to user
4. Select Approach
Read approach selection instructions and execute.
5. Report
- Create and ave the comprehensive report document:
{docs_dir}/{task_name}/feasibility/feasibility_report.md— see report format
- Use document-summarizer to create max 100-line summary
- Display the summary to the terminal
6. Complete
- TaskUpdate: mark completed
- TaskCreate:
Solution Design: <task-name>as next step - Display next phase command:
Next: /enterprise-dev-suite:solution-design -s {path-to-feasibility-report} - AskUserQuestion: proceed to next phase? (Yes: copy and run the command above / No: run later / Type Anything)
Constraints
- DRY: reuse existing code
- Docs dir:
.claude/claudeRes/docs(.claude is project-level one) - Document language: see conventions
Source
git clone https://github.com/arudita-zzz/app-dev-prompt-suite/blob/main/plugins/enterprise-dev-suite/skills/feasibility-study/SKILL.mdView on GitHub Overview
Reads a feature spec from .claude/claudeRes/scripts/feature_spec.md and performs a deep feasibility study that includes codebase analysis and web research. It also prototypes a PoC and documents findings to guide implementation decisions and next steps.
How This Skill Works
It starts by parsing optional research path (-r/--research) and initiating an interactive clarification with the user. It follows a staged workflow (Init, Clarify, Investigate, Propose, Select, Report, Complete), analyzes the codebase, conducts web research, and generates a feasibility report plus a concise summary for quick review.
When to Use It
- When starting a new feature with a formal spec
- When uncertain about technical constraints or integration complexity
- When evaluating multiple implementation approaches
- When a PoC is needed to validate feasibility
- When preparing documentation for management review
Quick Start
- Step 1: Initialize by reading the target spec and optional research path (-r/--research).
- Step 2: Engage the user to clarify the task name and constraints.
- Step 3: Run investigation, generate candidate approaches, create feasibility report, and prototype a PoC.
Best Practices
- Always read the feature spec at .claude/claudeRes/scripts/feature_spec.md
- Consult the senior engineer for any unclear requirements
- Reuse existing utilities and project conventions to DRY up work
- Document assumptions, risks, and decision criteria clearly
- Store outputs under .claude/claudeRes/docs and summarize with the document-summarizer
Example Use Cases
- Assessing feasibility of a new API integration against the spec
- Comparing frontend-first vs backend-first approaches for a feature
- Prototyping a PoC to validate data pipeline changes
- Evaluating security and compliance implications for a feature
- Estimating scope and risks for migrating a legacy module