菜单
Interactive installer for Rosetta. Guides users through selecting and installing skills and language-specific rules at user-level or project-level, verifies paths, and optionally optimizes installed files.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when designing memory or persistent context for an agent — deciding what to persist vs. recompute, how to retrieve only what's relevant, and how to summarize/compact history to fit the context window. Use when an agent forgets, bloats its context, or repeats work.
Instinct-based learning system that observes sessions via hooks, creates atomic instincts with confidence scoring, and evolves them into skills, commands, and agents.
Use when building or reviewing an evaluation for an LLM feature — assembling a representative test set, choosing pass criteria (exact match, programmatic checks, rubric, or LLM-as-judge), and catching regressions. Use when asking "how do I know this prompt or model change is better?"
Use when writing or reviewing a prompt for an LLM — making it specific, structured, and testable (clear task, role/context, examples, an explicit output contract, decomposition, room to reason) and avoiding the anti-patterns that make prompts brittle. Provider-agnostic.
Pattern for progressively refining context retrieval to solve the subagent context problem.
Use when creating or structuring a Claude Code plugin for this marketplace — the on-disk layout, the plugin.json manifest fields, the frontmatter for skills/commands/agents, and the authoring rules (no absolute paths, least-privilege tools, one coherent capability).
| name | configure-rosetta |
| description | Interactive installer for Rosetta. Guides users through selecting and installing skills and language-specific rules at user-level or project-level, verifies paths, and optionally optimizes installed files. |
Interactive step-by-step installation wizard for Rosetta. Uses AskUserQuestion throughout to guide selection and confirm choices.
User says "configure rosetta", "install rosetta", "setup rosetta", or asks to selectively install, verify, or optimize a Rosetta installation.
rm -rf /tmp/rosetta
git clone https://github.com/meaganewaller/rosetta /tmp/rosetta
Set ROSETTA_ROOT=/tmp/rosetta. If the clone fails, ask the user for a local path.
Ask: where to install — user-level (~/.claude/), project-level (.claude/), or both. Set TARGET accordingly and create $TARGET/skills and $TARGET/rules if they don't exist.
Ask which skill categories to install (multi-select): Framework & Language, Database, Workflow & Quality, or All. For each selected category, show the skills available and let the user deselect any.
Available skills are listed in the Rosetta repository under skills/. Read the directory rather than assuming a fixed list — the catalog changes.
For each selected skill:
cp -r $ROSETTA_ROOT/skills/<skill-name> $TARGET/skills/
Note: continuous-learning includes hooks and scripts beyond SKILL.md — copy the entire directory.
Ask which rule sets to install (multi-select): common (recommended), TypeScript, Python, Go, and any other language directories present in $ROSETTA_ROOT/rules/.
cp -r $ROSETTA_ROOT/rules/common/* $TARGET/rules/
cp -r $ROSETTA_ROOT/rules/<language>/* $TARGET/rules/
If the user selects language-specific rules without common rules, warn them — language rules extend common rules and may be incomplete without them.
Important: Rules install flat into $TARGET/rules/ — not in subdirectories. $TARGET/rules/coding-style.md is correct; $TARGET/rules/common/coding-style.md is not.
After installation:
grep -rn "~/.claude/" $TARGET/skills/ $TARGET/rules/
grep -rn "../common/" $TARGET/rules/
~/.claude/skills/ or ~/.claude/rules/ — check whether the referenced skill was also installed*-tdd → *-patterns) should have their dependencies installed too. continuous-learning references ~/.claude/homunculus/ — this is always user-level and is expected.For each issue: report the file, line, what's wrong, and what to do.
Ask whether to optimize installed skills, rules, both, or skip.
Skills: Read each SKILL.md, ask about the project's tech stack, suggest and apply removals of irrelevant sections, fix any path issues from Step 4.
Rules: Ask about preferences (coverage target, formatting tools, git workflow, security requirements) and edit rule files accordingly.
Only modify files in $TARGET/ — never touch $ROSETTA_ROOT/.
rm -rf /tmp/rosetta
Report: installation level and path, skills installed (count + names), rules installed (count by set), verification results, and any optimizations applied.