Skip to content

CLAUDE.md

Latest commit

 

History

History
72 lines (55 loc) · 4.67 KB

CLAUDE.md

File metadata and controls

72 lines (55 loc) · 4.67 KB
Raw

Claude Code Instructions For Promptify

Project Role

Promptify is a Claude Code skill repository. The single skill at skills/promptify/SKILL.md converts short developer intent into a structured task brief, shows the brief first, and asks whether to enter execution.

This repository is Markdown-only. Do not add a runtime service, npm CLI, package manager, web UI, database, telemetry, cloud sync, MCP server, slash commands, or non-Claude-Code adapters.

Claude Code Context

Load and reason from the repository root. The skill under skills/promptify/ depends on files under skills/promptify/shared/, which must stay inside the skill directory for Claude Code installation.

Key files:

  • skills/promptify/SKILL.md: The Promptify skill.
  • skills/promptify/shared/brief-standard.md: Generated brief fields, modes, and language rules.
  • skills/promptify/shared/glossary.md: Canonical Promptify terminology.
  • skills/promptify/shared/task-routing.md: Routing cues and priority.
  • skills/promptify/shared/safety.md: Safety levels, high-risk signals, and confirmation behavior.
  • skills/promptify/shared/context-discovery.md: Minimal-context exploration rules.
  • skills/promptify/shared/evolution-loop.md: SIA-style generational loop contract for the evolve task type (role mapping, artifact layout, workflow script contract, guardrails).
  • skills/promptify/shared/template-authoring.md: Rules for creating and editing compact templates.
  • skills/promptify/shared/templates/*.md: Core workflow templates (task, bugfix, feature, prototype, data analysis, evolve, refactor, test, review, docs, plan, PRD, goal, handoff).
  • docs/adr/: Architectural decision records for product boundaries and durable design decisions.
  • docs/out-of-scope/: Rationale for product directions Promptify intentionally does not support.

Editing Rules

  • Put platform-neutral behavior in skills/promptify/shared/.
  • Put skill-specific routing and behavior only in skills/promptify/SKILL.md.
  • Treat all shared/... references in the skill as relative to skills/promptify/; keep shared/ inside the promptify skill directory.
  • Use skills/promptify/shared/glossary.md terminology consistently when editing docs, templates, or skill behavior.
  • Use skills/promptify/shared/template-authoring.md when changing skills/promptify/shared/templates/ or the brief contract.
  • Match generated brief language to the user's input language by default while preserving technical identifiers, commands, paths, package names, and framework names.
  • The promptify skill is the only entry point. There are no slash commands, no host abstractions, no install/update/uninstall CLIs.
  • Record durable product-boundary decisions in docs/adr/; record rejected recurring feature directions in docs/out-of-scope/.
  • Preserve mode semantics inside the skill:
    • default prompt-first: output the brief, then ask whether to enter execution.
    • prompt-only: output the brief and stop.
    • review-only: findings first; do not edit unless explicitly requested.
    • plan-only: produce the plan and stop.
    • prd-only: produce the PRD and stop; do not publish to an issue tracker.
    • goal: produce only the goal block from skills/promptify/shared/templates/goal.md and stop.
    • handoff: produce only the handoff prompt from skills/promptify/shared/templates/handoff.md and stop.
    • analysis-first: auto-enforced on high-risk signals.

Safety Rules

  • High-risk signals include deletion, migration, payment, permission, auth, security, production, mass update, rewrite, and purge.
  • High-risk or destructive work must start with analysis and require explicit confirmation before edits.
  • Preserve unrelated user changes.
  • For review-only and plan-only flows, do not imply files were changed unless edits were explicitly requested later.

Verification

After changing the SKILL.md, shared templates, README, AGENTS.md, or this file, run the relevant checks:

rg -n "目标：|模式：|上下文：|要求：" skills/promptify/shared/templates -g '!prd.md'
rg -n "问题陈述：|解决方案：|用户故事：|实现决策：|测试决策：|非目标：|补充说明：" skills/promptify/shared/templates/prd.md
rg -n "analysis-first|prompt-only|review-only|plan-only|prd-only|shared/templates" skills
rg -n "T[B]D|T[O]DO|implement late[r]|fill in detail[s]" skills/promptify/shared skills README.md README.zh-CN.md AGENTS.md CLAUDE.md
git diff --check HEAD

The unfinished-marker scan should produce no output. If a check is not relevant to the files changed, say why in the final response.

Git Notes

  • Use a feature branch or isolated worktree for multi-step changes.
  • Keep commits scoped by concern.
  • Do not revert unrelated user changes.