ai-virtuoso
claude2 AI engineering skills: LLM evals and LLM observability
43 skills
27
claude
loop-virtuoso
Claude Code only. A backlog-driven delivery loop that feeds the agent one small item at a time and accepts it only when verify commands pass and a write-boundary guard approves. Also opt-in guard hooks, a settings hardener, a setup audit (doctor), setup evals against vanilla Claude Code, and starter templates. Ships hooks, scripts, commands, and 2 skills
/
Plugin loop-virtuoso from marketplace code-virtuoso in krzysztofsurdy/code-virtuoso. Commands read from the repo's manifest — 16 days ago.
This plugin runs code on your machine — its hooks fire automatically on events like tool use. The exact commands are listed below; read them before installing.
What it installs
Skills (3)
backlog-compilerConverts a PRD, ticket, or feature description into a .code-virtuoso/loop/backlog.json plan sized and ordered for autonomous loop execution. Use when preparing work for the iteration-loop skill, turning a written ticket into a loop-ready backlog, or when asked to turn a spec into a backlog.
iteration-loopExplains how the autonomous delivery loop runs over .code-virtuoso/loop/backlog.json, its state and stop reasons, the opt-in loop.guard hooks, and the harden, doctor and eval-setup commands. Use after backlog-compiler produces a backlog, or when asked how the loop or its guardrails work.
self-improveHow loop-virtuoso learns across runs. The self-improvement workspace, the learning schema, the durability capture gate, how the loop reads learnings back, and the /improve review that turns captured signals into commit-ready proposals. Use when running /loop-virtuoso:improve, when the loop reads or
Slash commands (10)
/loop-virtuoso:cancelCancel the active delivery loop; backlog state is left untouched
/loop-virtuoso:doctorRead-only audit of your Claude Code setup: duplicate skills and agents, skill listing size, CLAUDE.md references, plaintext MCP secrets, settings gaps, orphan plugin caches
/loop-virtuoso:eval-setupMeasure your Claude Code setup against vanilla Claude Code on a fixed task suite (costs money)
/loop-virtuoso:hardenMerge a guardrail baseline into Claude Code settings: dry-run diff first, apply only after approval
/loop-virtuoso:helpExplain what loop-virtuoso does, its two modes, and its commands
/loop-virtuoso:improveReview captured signals and the learnings store, optionally research current best practice, refresh the store, and report what is worth committing to a repo as a ready-to-apply diff. Human-gated: it never commits or wires anything itself.
/loop-virtuoso:init-setupScaffold the self-improvement workspace in the plugin's persistent space (survives updates): the learnings store and on-the-go signal capture. The capture opt-in is decided here.
/loop-virtuoso:scaffoldCopy a starter template into the repo: devcontainer, issue-to-PR workflow, or pilot backlog
/loop-virtuoso:startStart the session-supervised delivery loop (Mode A) over a backlog
/loop-virtuoso:statusReport loop progress: counts, iteration, stalls, log tail
Hooks (5)
- PreToolUse
bash "${CLAUDE_PLUGIN_ROOT}/hooks/guard-protected-paths.sh"
- PreToolUse
bash "${CLAUDE_PLUGIN_ROOT}/hooks/guard-secrets.sh"
- PreToolUse
bash "${CLAUDE_PLUGIN_ROOT}/hooks/guard-tests.sh"
- Stop
bash "${CLAUDE_PLUGIN_ROOT}/hooks/stop-hook.sh"
- Stop
bash "${CLAUDE_PLUGIN_ROOT}/hooks/guard-stop.sh"
More from krzysztofsurdy/code-virtuoso
All 7 plugins →frameworks-virtuoso
claude6 framework skills: Symfony upgrade, PHP upgrade, Composer dependencies, Godot and Godot upgrade, and Vercel AI SDK
43 skills
27
knowledge-virtuoso
claude6 knowledge skills: accessibility, context engineering, debugging, docs lookup, performance, and security
43 skills
27
playbooks-virtuoso
claude8 playbook skills: spec-driven development, TDD, branch finishing, PR feedback, handoff, hooks and guardrails, game feature loop, and security research
43 skills
27
tools-virtuoso
claude13 tool skills: rules writer, ticket writer, agent creator, skill writer, skill measure, plugin creator, research, brainstorming, communication style, PR message writer, report writer, stakeholder update writer, and UI review
43 skills
27
agents-virtuoso
claude20 agents (specialists and roles), 3 team workflows, and the verification skill, plus headless wrapper skills for CI review, test gaps, migration planning, dependency audit, and refactor scans
20 agents
27
Similar plugins
ralph-loop
claudeInteractive self-referential AI loops for iterative development. Claude works on the same task repeatedly, seeing its previous work, until completion.
3 commands · hooks
anthropics/claude-code 155,476
ruflo-iot-cognitum
claudeIoT device lifecycle, telemetry anomaly detection, fleet management, and witness chain verification for Cognitum Seed hardware
5 skills · 1 command · 4 agents
ruvnet/ruflo 81,589
hunter
claudecodexFind and verify professional email addresses, search contacts by domain, and enrich company data -- directly in Claude.
13 skills · 1 MCP server
anthropics/claude-plugins-official 47,076
operating-kit
claudePortable operating method: session lifecycle agents, pre-ship code review, deploy with live verification and state doc update, and production log health check
5 agents
wshobson/agents 42,461
Carbone
claudeOfficial Carbone skill — complete templating language reference covering tags, loops, conditions, formatters, aggregators, and all output formats (DOCX, XLSX, PPTX, ODT, HTML, Markdown, PDF)
1 skill
anthropics/knowledge-work-plugins 35,601
agent-launcher-skills
claudeBuild, launch, grade, and schedule Claude Managed Agents (CMA) in your own Anthropic account — a plugin re-implementation of Anthropic's launch-your-agent reference skill (Apache-2.0; independent, not a fork). Every session starts with a goal (./my-agent/goal.json, opt-in SessionStart hook, driven by /cs-goal); loop_compiler.py compiles it into a bounded grade->iterate loop (max_iterations 1..20), a recurring cron scheduled-deployment loop, or a single-pass interview->stage->launch workflow. 6 skills: orchestrator (context: fork goal router) + interview + stage-launch + grade-iterate + run-without-you + wrap-up. 18 stdlib-only deterministic scaffolders (no network/API calls; live launches emitted as BYOK curl that never prints the key), 4 agents, 8 commands, hooks, 5 references, 4 assets. Validators enforce CMA limits. Distinct from engineering/agent-harness (generic domain loop) and write-a-skill (authors Claude Code skills, not CMAs).
6 skills · 8 commands · 4 agents · hooks
alirezarezvani/claude-skills 29,725
Maintain loop-virtuoso?
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.
claude & codex plugins7 plugins