AgentSkillsCN

setup

配置 what-did-you-do 插件:链接至 i-did-a-thing 数据,设置面试偏好,并初始化会话跟踪。

SKILL.md
--- frontmatter
name: setup
description: "Configure what-did-you-do plugin: link to i-did-a-thing data, set interview preferences, and initialize session tracking"
disable-model-invocation: true
allowed-tools: Read, Write, Edit, Bash, Glob, Grep, AskUserQuestion
argument-hint: "[reconfigure]"

Set Up what-did-you-do

Configure the plugin to use your i-did-a-thing arsenal for interview preparation, practice sessions, and company-specific mock interviews.

Steps

1. Check for Existing Configuration

Read .claude/what-did-you-do.local.md. If it exists, tell the user:

Found existing configuration. I'll walk you through updating it — your current settings will be shown as defaults.

2. Verify i-did-a-thing Integration

Read .claude/i-did-a-thing.local.md to get the user's things_path. If missing:

This plugin works best with i-did-a-thing. Please run /i-did-a-thing:setup first to set up your accomplishment tracking.

You can still use what-did-you-do without it, but you won't get arsenal-powered feedback.

If found, confirm:

Found your i-did-a-thing config. Your arsenal at <things_path> will power your interview feedback.

Extract from the i-did-a-thing config:

  • things_path — where accomplishments live
  • current_role — for level-appropriate questions
  • target_roles — for gap-targeted practice
  • building_skills — for question weighting
  • aspirational_skills — for bonus weighting in question selection

3. Gather Interview Preferences

Use AskUserQuestion to ask:

How detailed should follow-up questions be during practice?

  • concise — no follow-ups, keep it moving
  • detailed — first-level follow-ups to test depth
  • coaching — deep follow-ups with probing questions and action items

Then ask:

Default interview stage to practice?

  • phone-screen — 30-min behavioral/culture fit
  • technical — coding and system design
  • onsite — full-day multi-round simulation
  • bar-raiser — cross-functional deep dive
  • no-default — ask me each time

4. Gather Trusted Sources for Question Updates

Use AskUserQuestion to ask:

Do you want to enable question bank updates from external sources?

  • Yes — I'll provide trusted domains
  • No — only use the built-in question bank

If yes, ask:

Trusted domains for question sources (comma-separated, e.g., leetcode.com, teamblind.com, levels.fyi)

And:

Trusted URLs (specific pages, comma-separated — optional)

5. Initialize Interview Prep Directory

Create the directory structure at <things_path>/interview-prep/:

code
<things_path>/interview-prep/
├── sessions/
├── progress.md
├── companies/
└── question-overrides/
    └── custom.yaml

Run via Bash:

bash
mkdir -p <things_path>/interview-prep/sessions
mkdir -p <things_path>/interview-prep/companies
mkdir -p <things_path>/interview-prep/question-overrides

6. Create Progress Dashboard

Write <things_path>/interview-prep/progress.md:

markdown
---
title: "Interview Readiness Dashboard"
total_sessions: 0
last_session: null
dimensions:
  specificity: null
  structure: null
  impact: null
  relevance: null
  self_advocacy: null
strongest_categories: []
weakest_categories: []
last_updated: <current_date>
---

# Interview Readiness

This dashboard is automatically maintained by what-did-you-do.

## Overall Readiness

_No sessions yet. Run `/what-did-you-do:practice` to start._

## Dimension Scores

| Dimension | Average | Trend | Last Session |
|-----------|---------|-------|-------------|
| Specificity | — | — | — |
| Structure | — | — | — |
| Impact | — | — | — |
| Relevance | — | — | — |
| Self-Advocacy | — | — | — |

## Category Breakdown

_Categories will appear as you practice different question types._

## Recent Sessions

_Sessions will be listed here._

7. Create Custom Questions Starter

Write <things_path>/interview-prep/question-overrides/custom.yaml:

yaml
# Custom interview questions
# Add your own questions here following the schema:
#
# - id: custom-001
#   text: "Your question here"
#   category: behavioral
#   subcategory: custom
#   skills_tested: [skill-1, skill-2]
#   level: [mid, senior]
#   stages: [phone-screen, onsite]
#   interviewer_types: [engineering-manager]
#   follow_ups:
#     - text: "Follow-up question"
#       depth: 2
#   difficulty: 3
#   expected_format: narrative
#   time_budget_minutes: 5
#   red_flags: ["vague answer"]
#   green_flags: ["specific metrics"]

questions: []

8. Write Configuration

Write .claude/what-did-you-do.local.md:

yaml
---
things_path: "<from i-did-a-thing config>"
follow_up_depth: "<concise|detailed|coaching>"
default_stage: "<stage or no-default>"
trusted_sources:
  domains:
    - "<domain>"
  urls:
    - "<url>"
current_role: "<from i-did-a-thing config>"
target_roles:
  - "<from i-did-a-thing config>"
building_skills:
  - "<from i-did-a-thing config>"
aspirational_skills:
  - "<from i-did-a-thing config>"
last_updated: <current_date>
---

# what-did-you-do Configuration

Generated by `/what-did-you-do:setup`. Run it again to reconfigure.

These settings are used by all what-did-you-do skills:
- `practice` — single-question drill with persona-driven feedback
- `mock` — full interview round simulation
- `review` — readiness assessment across dimensions
- `prep-for` — company-specific preparation plan
- `update-questions` — add questions from trusted sources

9. Confirm Setup

Tell the user:

Your what-did-you-do setup is complete!

  • Arsenal: <things_path>
  • Follow-up depth: <depth>
  • Default stage: <stage>
  • Trusted sources: <count or "built-in only">

Quick start:

  • /what-did-you-do:practice — Drill a single question with coached feedback
  • /what-did-you-do:mock — Simulate a full interview round
  • /what-did-you-do:prep-for — Prepare for a specific company
  • /what-did-you-do:review — Check your readiness across dimensions