bootstrap

📁 jwilger/agent-skills 📅 1 day ago
0
总安装量
1
周安装量
安装命令
npx skills add https://github.com/jwilger/agent-skills --skill bootstrap

Agent 安装分布

mcpjam 1
claude-code 1
replit 1
junie 1
zencoder 1

Skill 文档

Bootstrap

Value: Communication — help the developer and the agent establish a shared understanding of what workflow is appropriate before any work begins.

Purpose

Entry point for developers who have no SDLC skills installed or do not know where to start. Detects the project environment, asks two questions, and recommends specific skills the developer can install. Never silently installs anything.

Practices

Detect the Environment

Before asking questions, gather project context silently:

!test -f package.json && echo "js" || true
!test -f Cargo.toml && echo "rust" || true
!test -f pyproject.toml && echo "python" || true
!test -f go.mod && echo "go" || true
!test -f mix.exs && echo "elixir" || true
!git rev-parse --is-inside-work-tree 2>/dev/null && echo "git" || true

Use detected languages and tools to tailor recommendations.

Ask Two Questions

Question 1: What are you trying to do?

  • “Start a new project” — recommend Understand + Decide + Build phases
  • “Add a feature or fix a bug” — recommend Build + Ship phases
  • “Review or ship existing work” — recommend Ship phase only
  • “Set up team workflow” — recommend all phases plus orchestration

Question 2: How much process structure do you want?

  • “Minimal — just help me write good code” — recommend tdd-cycle, debugging-protocol
  • “Standard — tests, reviews, and architecture” — recommend core + ship skills
  • “Full — event modeling, domain-driven, the works” — recommend all skills

Recommend Skills by Phase

Present recommendations grouped by phase. Show install commands. Let the developer choose which to install.

Understand (discovery and requirements):

  • event-modeling — event-driven design, swimlanes, GWT scenarios

Decide (architecture and domain):

  • architecture-decisions — lightweight ADR governance
  • domain-modeling — parse-don’t-validate, type-driven design

Build (implementation):

  • tdd-cycle — red/green/refactor with domain review checkpoints
  • debugging-protocol — systematic 4-phase debugging
  • user-input-protocol — structured checkpoints when input is needed

Ship (review and delivery):

  • code-review — three-stage review protocol
  • mutation-testing — test quality verification

Workflow (coordination):

  • orchestration — multi-agent delegation patterns
  • task-management — work breakdown and state tracking
  • memory-protocol — cross-session knowledge persistence

Install command format:

npx skills add jwilger/agent-skills --skill <skill-name>

Let the Developer Decide

After presenting recommendations, wait for the developer to confirm which skills to install. Do not install anything without explicit confirmation. If the developer wants everything, provide a single combined command.

Enforcement Note

This skill is purely advisory. It recommends skills but cannot install them. The developer must run install commands themselves or confirm installation. For available enforcement plugins, see the Harness Plugin Availability table.

Verification

After applying this skill, verify:

  • Environment was detected before asking questions
  • No more than 2-3 questions were asked
  • Recommendations were grouped by phase
  • Install commands were shown for each recommended skill
  • Nothing was installed without developer confirmation

Dependencies

This skill works standalone. It recommends but does not require other skills. All recommended skills are independently installable.