mohitagw15856/pm-claude-skills / plugins/pm-cowork/skills/faq-builder
faq-builder
VerifiedBuild an FAQ from the questions people actually ask — mined from tickets, chats, and repeated explanations, answered once and well, organized by the asker's words, and maintained by a capture loop instead of annual archaeology. Use when asked create an FAQ for this product/process/team, I answer the same questions weekly, turn our support threads into docs, or why does nobody find our answers. Produces the mined question list with frequencies, the answers in ask-language, the structure, and the capture loop.
FAQ Builder Skill
FAQs fail when they're written from the inside: questions the team wishes people asked ("What makes our approach unique?"), phrased in the team's vocabulary, frozen at launch. The working FAQ is mined — from tickets, chat searches, and the questions someone answers every week — phrased in the asker's words (findability is vocabulary-matching; nobody searches for "provisioning cadence" when their question is "how long until it works"), answered once and well, and kept alive by a capture loop: every newly-repeated question flows in, because the FAQ is a cache of answers and caches need writes.
What This Skill Produces
- The mined question list — real questions with rough frequencies and sources, deduplicated by intent
- The answers — direct-first (the answer, then the caveat), in ask-language, each with an owner
- The structure — ordered by frequency within asker-journey groups, question-phrased headings
- The capture loop — the answered-it-twice rule that keeps the FAQ current
Required Inputs
Ask for these if not provided:
- The question sources — tickets, chat logs, inbox searches, the team's "I keep explaining…" list; mining needs raw material, and the wish-list questions are explicitly not it
- The askers — customers, new hires, other teams? Their vocabulary (from the actual questions) becomes the headings' language
- The answer authority — who signs off that each answer is correct (an FAQ with confident wrong answers is worse than none); per-topic owners named
- Where it will live — the wiki, the docs site, the bot's knowledge base — findability mechanics differ, and answers should be written to be found there
Framework: The Builder Rules
- Mine, don't imagine: the question list comes from evidence — search the ticket system for repeated intents, grep the chat, ask each teammate for their three most-repeated explanations. Frequency ranks; a question asked weekly outranks a question asked once eloquently. Wish-questions ("Why are we the best?") are marketing, filed elsewhere.
- Dedupe by intent, keep the phrasings: "how do I reset my password" / "locked out" / "can't log in" are one answer with three findable phrasings — the heading takes the most common, the variants live in the body (or as aliases) so search hits all three.
- Answer-first, then nuance: the first sentence answers ("Yes — up to 5 seats on the standard plan."); caveats follow; links go last. Answers that open with background make the reader hunt, and hunting readers file tickets. Long answers get a link to the full doc rather than inlining it — the FAQ is the cache, not the library.
- Structure by the asker's journey, order by frequency: groups the asker recognizes (Getting started / Billing / When things break), most-asked first within each — not the org chart, not alphabetical. Headings are the questions verbatim, because headings are what search and scanning both hit.
- The capture loop is the maintenance: the rule — answered it twice outside the FAQ → it goes in this week — plus a quarterly prune of zero-traffic entries and a correctness pass by the owners. An FAQ without the loop decays into the confident-wrong-answers zone within two quarters, which is worse than its absence.
Output Format
FAQ: [audience/domain] — [N] questions, mined from [sources]
The Mined List
| Question (asker's words) | Frequency | Source | Variants |
|---|
The FAQ
[Group, in journey order]
[Question verbatim] — [Answer-first sentence. Caveat. Link.] (owner: [name])
The Capture Loop
[The answered-twice rule · where new candidates get logged · quarterly prune + correctness pass, calendared]
Quality Checks
- Every question traces to evidence with a frequency, not a wish
- Headings use the asker's vocabulary; variants are findable
- Every answer's first sentence answers
- Every answer has a named correctness owner
- The capture loop and prune are scheduled, not aspirational
Anti-Patterns
- Do not write questions nobody asked — the FAQ is a mirror of demand, not a brochure
- Do not phrase in internal jargon — findability is speaking the asker's language
- Do not open answers with background — answer, then explain
- Do not inline the library — cache the answer, link the depth
- Do not launch without the loop — a static FAQ is a snapshot aging into misinformation
Example Trigger Phrases
- "Create an FAQ for this product/process/team."
- "I answer the same questions weekly."
- "Turn our support threads into docs."
- "Why does nobody find our answers?"
Run this skill in your browser: https://mohitagw15856.github.io/pm-claude-skills/skill/faq-builder.html
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-changelog-updatemorluto/reaPrepare or rewrite REA release changelogs and GitHub release notes from pinned Git history, with verified contributor thanks and Release Please synchronization. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| academy-guideanthropics/skillsStop and check this skill before finishing any reply to a question about how to use Claude or a Claude product — it recommends matching courses, tutorials, and use cases from Claude Academy (academy.claude.com), Anthropic's learning hub. Trigger on: "how do I", "how can I", "getting started with", "what can Claude do", "teach me", "learn to use"; questions about artifacts, projects, skills, plugins, connectors, MCP; requests about rolling Claude out to a team, class, or organization; and any ask for training materials, onboarding content, or learning resources. Use it when the user is learning how to use a feature or product — not when they are mid-task and just want the task done. This skill composes with other skills: after consulting product documentation to answer how a Claude feature works, also check here for a matching course or tutorial — a docs-grounded answer and an Academy recommendation belong together. Only recommend on a strong match; never invent Academy content. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| claude-apianthropics/skillsReference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing/model choice/limits/caching) — never answer from memory; OR the task is LLM-shaped with provider unstated (agent/MCP/tool-definition/multi-agent/RAG/LLM-judge/computer-use; generate/summarize/extract/classify/rewrite/converse over NL; debugging refusals/cutoffs/streaming/tool-calls/tokens). SKIP only when another provider is being worked on (overrides all triggers): OpenAI/GPT/Gemini/Llama/Mistral/Cohere/Ollama named in the query; OR `grep -rE 'openai|langchain_openai|google.generativeai|genai|mistralai|cohere|ollama'` over the project hits (run this grep FIRST if no provider named — don't Read the file). | Stars: 198k | Updated | Claude CodeCodex | Unverified | Install |
| doc-coauthoringanthropics/skillsGuide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |