Interactive qa
Custom Claude Code skills for session handoff, verification, planning, research, audits, and decision-gathering.
npx -y skills add giovicordova/skills-gc --skill interactive-qaAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Interactive Q&A that collects the user's decisions when agents present multiple options or when multiple agents disagree. Produces a short plain-language brief the user can hand to Plan Mode, a specialized agent, or keep in context. Use this skill whenever any agent (or a board-meeting, brainstorming, or pre-plan session) comes back with 2+ reasonable options, trade-offs, or conflicting recommendations that need a human call — the user is not a developer, so options must be in non-technical plain English with the agent's recommendation flagged and any disagreement between agents surfaced. Also trigger on 'ask me', 'Q&A me', 'let me decide', 'collect my decisions', 'interactive Q&A', 'I need to decide', or any moment the current step is blocked waiting on a human decision. Do not guess the user's preference — ask.
SKILL.md
5.2 KB, ~1.0k tokens by cl100k_base, as published. Nobody here has run it
Interactive Q&A
Collect the user's decisions through structured Q&A. Turn agent-level trade-offs into plain-language choices the user can answer in seconds.
When this skill runs
- Board meetings — agents have weighed in, views conflict, the user settles it.
- Brainstorming — ideas are branching, the user's taste picks the branch.
- Pre-plan briefs — before Plan Mode, gather the decisions the plan depends on.
- Agent stalls — any agent is blocked on a choice only the user can make.
- Multi-agent disagreement — two or more agents argue for different options on the same subject.
If there are 2+ reasonable options on the table, do not guess. Ask.
Principles
- Plain language. The user is not a developer. Describe options like you'd describe them to a smart friend over coffee. No jargon, no acronyms, no framework names unless strictly necessary.
- Brief. 1–5 word labels. One-sentence descriptions that name the trade-off in practice.
- Flag the recommendation. Put the recommended option first and append "(Recommended)" to the label.
- Surface disagreement. When multiple agents disagree, make each agent's preferred option a distinct choice, name the agent in the description ("Marketing says…", "Business says…"), and include at least one compromise option.
- Don't infer. If an option is missing, let the user type "Other". Don't fill gaps silently.
The three steps
Step 1 — Ask about handoff first
Before any decision questions, ask one AskUserQuestion: where does the brief go when we're done? This shapes the rest of the flow.
Default options (adapt to project):
- Keep in context (Recommended) — brief stays inline; the next agent or Claude reads it from the conversation.
- Feed to a specific agent — if the project has a specialist ready (e.g. web-developer, editorial, marketing). Name the agent.
- Enter Plan Mode — hand the brief straight into Plan Mode as the starting spec.
- Save to file — write
BRIEF.md(or a topic-specific name) in the project root.
If the handoff is already clear from conversation ("then enter plan mode"), skip this call and confirm inline in one sentence.
Step 2 — Ask the decisions
Use AskUserQuestion, up to 4 questions per call. Batch thematically — don't mix unrelated decisions in one call.
For each question:
- One clear question ending in a question mark.
- 2–4 options. More than 4 means the question is too broad — split it.
- Recommended option first, with "(Recommended)" suffix.
- Option labels: 1–5 words of plain English.
- Option descriptions: one sentence describing what this choice means in practice and the main trade-off.
- multiSelect: true only when choices genuinely aren't mutually exclusive.
When agents disagree, format options like this:
Marketing's pick (Recommended) — bold headline, one CTA. Best for conversion. Least visual polish. Business's pick — lead with logos and reviews. Builds enterprise trust. Slower story. Editorial's pick — image-led with short tagline. Feels premium, on-brand. Weakest at conversion. Compromise — bold headline with logos below. Keeps conversion priority, addresses trust. Middle ground.
The user sees at a glance who's arguing for what — and why.
Step 3 — Produce the brief and hand off
Assemble a short brief. One line per decision. This is a brief, not a report.
## Decisions
- [Topic]: [chosen option] — [one-line note if the user added context]
- [Topic]: [chosen option]
- …
## Handoff
[One line: "Stays in context" / "Entering Plan Mode" / "Handing to web-developer agent" / "Saved to BRIEF.md"]
Then hand off based on Step 1:
- Keep in context — post the brief inline. Stop. The next turn picks it up.
- Feed to agent — dispatch the named agent with the brief as the input prompt.
- Enter Plan Mode — call EnterPlanMode using the brief as the starting spec.
- Save to file — Write the brief to
BRIEF.md(or{topic}-brief.md) at project root. Stop.
What this skill does not do
- Does not make decisions for the user. When in doubt, ask.
- Does not pad the flow with confirmations ("Got it!", "Great choice!"). Capture, brief, hand off.
- Does not write long reports. The brief is a handoff artefact, not a document.
- Does not re-ask the user things already decided earlier in the conversation.
Gives 0 of the 12 instructions most plan spec skills give in ~1.0k tokens
Counted across 1,100 of the 1,860 authors here whose files we hold, read 2026-08-06
- ask one question at a timein 46 of 1100, across 38 files
- Break plans into vertical slicesin 28 of 1100, across 10 files
- Publish issues in dependency orderin 27 of 1100, across 9 files
- Iterate until user approves the breakdownin 24 of 1100, across 6 files
- Explore the repository to understand the codebase statein 24 of 1100, across 7 files
- Use domain glossary vocabularyin 23 of 1100, across 5 files
- Apply correct triage labels to published issuesin 23 of 1100, across 5 files
- Write failing tests before implementation codein 23 of 1100, across 18 files
- Prefer AFK slices over HITLin 22 of 1100, across 7 files
- ask clarifying questions until requirements are concretein 21 of 1100, across 13 files
- Respect existing architecture decision recordsin 20 of 1100, across 5 files
- write a specification before writing any codein 20 of 1100, across 12 files
Said here and by no other author read
- present all options in plain language
- append recommended suffix to suggested options
- keep option labels under five words
- write one sentence option descriptions
- name the disagreeing agent in descriptions
- include a compromise option for disagreements
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.