project-readmelisted

# Project README Use this skill to build README files from repository facts. README structure is selected by the primary reader and their first successful action, not by language ecosystem alone. ## Workflow 1. Inspect the repository before writing: ```bash SCRIPT=$(find ~ -maxdepth 8 -name "inspect.py" -path "*/project-readme/scripts/*" 2>/dev/null | head -1) [ -n "$SCRIPT" ] && python3 "$SCRIPT" . || echo "Script not found; inspect manifests and repo facts manually." ``` 2. Use the inspector output: - `ecosystem` controls install commands, package metadata, badges, dev commands, and language-specific examples. - `readme_type` controls structure, first screen, section order, and reader journey. - `readme_type_candidates` is the ambiguity set. If `readme_type` is `unknown` and candidates exist, ask one clarifying question based on the candidate first actions. If work must proceed without an answer, choose the type whose first successful action is best supported by repo facts. - `agent_skills` lists detected skill files and plugin manifests when `readme_type` is `agent-skill`; use it to build collection indexes and validation sections. 3. Load only the needed modules: - Ecosystem: `ecosystems/<ecosystem>.md`, or each ecosystem when `ecosystem` is `mixed`. - Type: `project-types/<readme_type>.md`. - Shared ordering: `references/section-order.md`. - Badges: `references/badge-patterns.md`. - Validation rules: `references/validation.md`