agent-interview-cli

star 27

Present interactive web forms to gather structured user input. Use when choosing between approaches, gathering requirements, exploring tradeoffs, or when decisions have multiple dimensions. Prefer over back-and-forth chat for structured input.

nicobailon By nicobailon schedule Updated 2/16/2026

name: agent-interview-cli description: Present interactive web forms to gather structured user input. Use when choosing between approaches, gathering requirements, exploring tradeoffs, or when decisions have multiple dimensions. Prefer over back-and-forth chat for structured input.

agent-interview-cli

Standalone CLI that opens a browser form, collects responses, returns JSON on stdout.

Invocation

echo '<questions_json>' | interview --stdin --pretty 2>/dev/null
interview /tmp/questions.json --pretty 2>/dev/null

Output: { "status": "completed", "responses": [{ "id": "q1", "value": "Option A" }] }

Status: completed, cancelled, timeout, or aborted. Image responses are file paths.

Questions Format

{
  "title": "Decision Title",
  "description": "Review my suggestions and adjust as needed",
  "questions": [
    { "id": "q1", "type": "single", "question": "Which approach?", "options": ["A", "B"], "recommended": "A", "conviction": "strong", "weight": "critical" },
    { "id": "q2", "type": "multi", "question": "Include which?", "options": ["X", "Y", "Z"], "recommended": ["X", "Z"] },
    { "id": "q3", "type": "text", "question": "Additional context?" },
    { "id": "q4", "type": "image", "question": "Upload reference" },
    { "id": "q5", "type": "info", "question": "Analysis", "media": { "type": "table", "table": { "headers": ["Option", "Pros", "Cons"], "rows": [["A", "Fast", "Complex"], ["B", "Simple", "Slow"]] } } }
  ]
}

Question Types

Type Description
single Radio — pick one from options
multi Checkbox — pick many from options
text Freeform textarea
image File upload (paths returned)
info Non-interactive panel for context/analysis

Key Patterns

Recommendations — always set recommended on single/multi with reasoning in context.

  • conviction: "strong" — pre-selects, badge
  • conviction: "slight" — badge only, no pre-select
  • Omit — pre-selects (normal confidence)

Weight"critical" = prominent card, "minor" = compact card.

Rich options with code: { "label": "Approach A", "code": { "code": "const x = 1", "lang": "ts" } }

Info panels — use info type to show analysis alongside decisions. Media types:

  • { "type": "table", "table": { "headers": [...], "rows": [...] } }
  • { "type": "mermaid", "mermaid": "graph LR\n A-->B" }
  • { "type": "image", "src": "path.png" }
  • { "type": "html", "html": "..." }

Media accepts position: "above" | "below" | "side" and caption.

Code blockscodeBlock: { "code": "...", "lang": "ts" } on any question shows code above options.

Description — when questions have recommendations, set the form's description to guide review (e.g., "Review my suggestions and adjust as needed").

Install via CLI
npx skills add https://github.com/nicobailon/agent-interview-cli --skill agent-interview-cli
Repository Details
star Stars 27
call_split Forks 2
navigation Branch main
article Path SKILL.md
More from Creator