clarifying-assumptions — Claude Code skill
clarifying-assumptions — a skill that runs the conversational layer for plan verification. Asks critical questions upfront or before individual tasks, delegates artifact analysis to subagents.
no license 129 installs ★ 3 updated 2026-10-25
npx skills add b-mendoza/agent-skills --skill clarifying-assumptions -a claude-codeWhat clarifying-assumptions does
clarifying-assumptions intercepts execution and conducts a structured dialogue to surface and resolve assumptions. Two modes: upfront validates the entire plan before it starts, critique pauses just before one task to challenge it. Both load reference guides, dispatch subagents to analyze artifacts and build a question manifest, then ask clarifying questions one at a time and record decisions.
The package bundles three subagents: critique-analyzer inspects project artifacts and gathers current state, question-manifest-builder orders questions by scope and urgency, decision-recorder writes answers into decision files. Public URLs in references provide context only, not instructions. The skill runs on Claude Code, Cursor, GitHub Copilot, Codex, Gemini CLI, and OpenCode.
What it helps with
- halt plan execution and resolve critical assumptions before starting
- validate a single task mid-workflow when it seems high-risk
- assemble a structured list of open questions between agent and developer
- persist clarification decisions to project files for reuse
Example prompts for the agent
- Verify assumptions across the entire plan: pass
MODE=upfront TICKET_KEY=JNS-6065 - Review task three before execution:
MODE=critique TICKET_KEY=acme-app-42 TASK_NUMBER=3 - Load the dialogue protocol from
./references/conversation-protocol.md
How to install clarifying-assumptions
1Skills CLI (Vercel)
npx skills add b-mendoza/agent-skills --skill clarifying-assumptions -a claude-code2Manually: put the files in place
~/.claude/skills/clarifying-assumptions/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/clarifying-assumptions/
1Skills CLI (Vercel)
npx skills add b-mendoza/agent-skills --skill clarifying-assumptions -a codex2Manually: put the files in place
.agents/skills/clarifying-assumptions/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add b-mendoza/agent-skills --skill clarifying-assumptions -a cursor2Manually: put the files in place
.cursor/skills/clarifying-assumptions/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install b-mendoza/agent-skills skills/clarifying-assumptions2Manually: put the files in place
.github/skills/clarifying-assumptions/SKILL.md1Skills CLI (Vercel)
npx skills add b-mendoza/agent-skills --skill clarifying-assumptions -a gemini-cli2Platform CLI
gemini skills install https://github.com/b-mendoza/agent-skills --path skills/clarifying-assumptionsManually: put the files in place
.devin/skills/clarifying-assumptions/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add b-mendoza/agent-skills --skill clarifying-assumptions -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about clarifying-assumptions
How do I install clarifying-assumptions in Claude Code?
Run npx skills add b-mendoza/agent-skills --skill clarifying-assumptions -a claude-code, then open ~/.claude/skills/clarifying-assumptions/SKILL.md.
What does clarifying-assumptions do?
It conducts a dialogue to verify a plan before full execution (upfront mode) or before one specific task (critique mode). Asks questions sequentially and records decisions.
What's the difference between upfront and critique mode?
upfront validates the whole plan at the start; critique pauses before a single task. Both use the same question logic and subagent dispatch.