Borrowing it
Nothing to install: this file belongs to OlliMakarova/mcp-metro. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/OlliMakarova/mcp-metro/master/.claude/skills/project-decision-interview/SKILL.mdgit clone --depth 1 https://github.com/OlliMakarova/mcp-metroWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/ollimakarova/mcp-metro/project-decision-interview)<a href="https://agentmods.dev/skills/ollimakarova/mcp-metro/project-decision-interview"><img src="https://agentmods.dev/badge/skills/ollimakarova/mcp-metro/project-decision-interview/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/ollimakarova/mcp-metro/project-decision-interview"><img src="https://agentmods.dev/badge/skills/ollimakarova/mcp-metro/project-decision-interview.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00076 | $0.01189 |
| Opus 5 | $0.00038 | $0.00594 |
| Sonnet 5 | $0.00015 | $0.00238 |
| Haiku 4.5 | $0.00008 | $0.00119 |
Grade A, and why
project-decision-interview scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Interactive project decision interview
Goal
Bring the task to a consistent and implementable target state. Do not stop at conversation: persist decisions, update the affected documents, and once the questions are over, carry out the implementation if it is part of the request.
Order of work
- Read the project rules, the state of the working copy, and the documents the user pointed to.
- Investigate the requirements, documentation, code, settings, data structures, checks, and change history that relate to the task. If the user named specific commits, study them and the resulting state of the code without fail.
- Compare the current design with the user's goal. Separate the facts that can be established from the project itself from the decisions that genuinely require the user's choice.
- Look for an existing working document with a plan or a decision register. If there is none and the user asked for decisions to be recorded, create a suitable document following the project rules.
- Build an internal list of the forks in the road, ordered by decreasing impact. Start with decisions that change the purpose of the system, the data model, permissions, data safety, or the user scenario.
- Ask one question, get the answer, immediately record the decision that was made, and only then move on to the next question.
- After the last answer, check that the target model is complete, carry out the implementation the request calls for, and confirm the result with checks.
How to ask questions
- Ask exactly one question per message.
- Ask a multiple-choice question with the
AskUserQuestiontool: two to four options, the recommended one first and marked "(Recommended)" in its label, each option carrying a description of its observable consequences. The free-form "Other" option is added by the tool itself. If the question does not reduce to a choice among options, ask it as plain text. - Before each question, briefly state which previous decision has been recorded and what it leads to.
- Explain an unclear term with a concrete example before repeating the question.
- Offer the recommended option and describe its observable behavior clearly.
- Where possible, phrase the question so that it can be answered with "yes" or "no".
- Do not ask a question whose answer can be reliably obtained from the code, the history, the settings, or the documentation.
- Do not push minor technical decisions onto the user. Make them yourself, following the project rules.
- If the user has not chosen an option, do not record it as accepted.
- If the user corrected a decision, replace the previous target state and delete the wording that contradicts it.
- If the user added a new thought, record its consequences and continue with the single most important fork in the road.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 90 lines · 76 tokens per session scan A cdef3d3fecea
project-decision-interview is a skill published in the GitHub repository OlliMakarova/mcp-metro (0 stars, last pushed 1mo ago), licensed MIT. It adds 76 tokens to every session and 1,189 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…