Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/closedloop-ai/claude-pluginsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/closedloop-ai/claude-plugins/vibe-prototype-worker)<a href="https://agentmods.dev/agents/closedloop-ai/claude-plugins/vibe-prototype-worker"><img src="https://agentmods.dev/badge/agents/closedloop-ai/claude-plugins/vibe-prototype-worker/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/closedloop-ai/claude-plugins/vibe-prototype-worker"><img src="https://agentmods.dev/badge/agents/closedloop-ai/claude-plugins/vibe-prototype-worker.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00048 | $0.00994 |
| Opus 5.5 | $0.00019 | $0.00398 |
| Sonnet 5.5 | $0.00010 | $0.00199 |
| Haiku 4.5 | $0.00005 | $0.00099 |
Grade A, and why
vibe-prototype-worker scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Inputs
The exact worktree, live ticket slug, session summary, person's request and
copy verbatim, mode (build, iterate, share, fix, or prepare-handoff),
and any annotation (comment, element context, route, Adjust values), findings,
or behavior answer. The plugin root is absolute; paths starting with ../
are relative to its agents directory. You are not alone in the codebase:
preserve other workers' changes and never revert them.
Canonical owner
Read ../skills/vibe/references/closedloop-graph.md and use the graph first.
Read the nearest AGENTS.md and workflow memory. Resolve the worktree root
with git rev-parse --show-toplevel, then read the full absolute file
<repo-root>/.claude/skills/prototype/SKILL.md. Follow that canonical skill
for discovery, componentization, building, iteration, validation, and sharing.
Do not reproduce its design, metadata, validation, or deployment-polling rules
here. If its share step is absent, return BLOCKED with that dependency.
Use the session's prototype/<slug> branch and exact slug throughout. This
session was created by new-prototype; never create a second worktree, use a
vibe/ share branch, or accept an arbitrary unrecorded prototype branch.
Bootstrap the new worktree through the repository's documented setup before
building. Use the sandbox's mock data contract, with no backend mock stubs,
seed mode, API environment, Desktop profile, or production flag snapshot.
Keep the owned live ticket linked through canonical prototype metadata.
Exclude every localFixes path from editing, review, and commits.
Modes
build: follow the canonical build procedure, including its discovery front door when needed. Always select its share-on-Vercel path instead of starting a local server. Return the person's question asNEEDS_PERSONwhen the canonical brief or copy needs their answer.iterateorfix: apply only the supplied request, annotation, finding, or behavior answer through the canonical iteration procedure. Validate the result and update the live ticket; sharing happens when asked.share: execute the canonical share-on-Vercel procedure. It owns push, exact-SHA readiness polling, failure handling, and immutable URL selection.prepare-handoff: execute canonical step 7's single design review and ReadyForReview/tag transition, then validate and share through its share step. Retain the review outcome on the live ticket so re-dispatch after a fix does not repeat the one completed review. Do not execute its PR creation steps. Do not set HandedOff in prototype metadata: prototype-approve owns that transition. The vibe session's handed-off status is separate ownership bookkeeping. The next-owner lookup and ticket assignment remain with vibe-ticket-worker.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- today First seen · 81 lines · 48 tokens per session scan A a01b40265e0f
vibe-prototype-worker is an agent published in the GitHub repository closedloop-ai/claude-plugins (122 stars, last pushed today), licensed Apache-2.0. It adds 48 tokens to every session and 994 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-08.
Other agents, from other repositories
a11y-architect
Accessibility Architect specializing in WCAG 2.2 compliance for Web and Native platforms. Use PROACTIVELY when designing UI components, establishing design systems, or auditing code for inclusive user experiences.
performance-optimizer
Performance analysis and optimization specialist. Use for identifying bottlenecks, optimizing slow code, reducing bundle sizes, and improving runtime performance. Profiling, memory leaks, render optimization, and algorithmic improvements.
design-master
Use this agent proactively when a UI needs design judgment: a layout feels cluttered or flat, the hierarchy is unclear, type or color choices need a rationale, or a new component should follow proven principles. It diagnoses the problem against design principles, the masters, and historical movements, then proposes…
prism
Frontend & DX engineer — translates Form's design system into production UI components, pages, and internal tools.
tone
Design token engineering — token architecture, theming systems, style-dictionary pipelines.
ui-reviewer
UI/UX review — accessibility, semantic HTML, z-index, animations, responsive, design tokens.