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.
npx skills add jamesyang124/agent-skills --skill agent-playtestgit clone --depth 1 https://github.com/jamesyang124/agent-skillsWrote 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/skills/jamesyang124/agent-skills/agent-playtest)<a href="https://agentmods.dev/skills/jamesyang124/agent-skills/agent-playtest"><img src="https://agentmods.dev/badge/skills/jamesyang124/agent-skills/agent-playtest/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/skills/jamesyang124/agent-skills/agent-playtest"><img src="https://agentmods.dev/badge/skills/jamesyang124/agent-skills/agent-playtest.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 | $0.00187 | $0.02251 |
| Opus 5.5 | $0.00075 | $0.00900 |
| Sonnet 5.5 | $0.00037 | $0.00450 |
| Haiku 4.5 | $0.00019 | $0.00225 |
Grade A, and why
agent-playtest 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 20d ago.
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 — 127 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent playtest: closed-loop play → screenshot evidence → described, audited data
Many catalog entries (rooms, games, worlds, apps) have publisher text that is wrong, empty, or templated. This skill gets an observed description: an agent actually opens each item, plays it, records what it saw, and writes a description in which every claim cites a screenshot. The output stays separate from the publisher's declared data and never overwrites it. Disagreements are recorded and handed to whoever owns the catalog.
Every item goes through this pipeline:
plan → dispatch N player subagents → play (closed loop) → audit screenshots → validate → spot-check → report
↘ blockers log (anything that could not be done: record it and continue) ↙
Roles
| Role | Who | Does |
|---|---|---|
| Orchestrator | the calling session | writes the plan and brief, dispatches players, audits, validates, samples the spot-check, reports, keeps the blockers log |
| Player | 1 subagent per slice of items (default ≤ 4 in parallel) | plays its items in order with one browser session, writes playtest.json, shots/, description.json |
| Human | the person running it (optional) | spot-checks descriptions; plays hard items in human-demo mode |
Evidence contract
Read references/evidence-review.md before capture or review: content-only screenshots, coverage by content type, declared/observed separation, timing limits, model roles and portable batches. It takes precedence over historical capture/report examples. Keep database schemas unchanged unless separately authorized.
Before you start (orchestrator)
- Pick a browser adapter. Read
references/browser-adapters.md. It lists the capabilities a player needs and how ego-browser, Playwright and Chrome DevTools provide each one. Record the choice in the campaign entry. - Pick or write a site profile. Site-specific rules (URLs, metadata source, paywalls, device gates, embed
origins) go in
profiles/<site>.md, never in the generic protocol.profiles/viverse.mdis the worked example. - Set up the workspace (layout in
references/data-schema.md). Then add acampaigns.jsonentry and writearchive/<campaign>/plan.json(templates/plan.example.json). - Write the brief from
templates/agent_brief.md, filling in the adapter, profile, paths, and any rules specific to this campaign.
What ships with it
32 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- assets/review/fonts/inter-latin-wght-normal.woff2 47 KB
- assets/review/licenses/daisyui-LICENSE.txt 1.0 KB
- assets/review/licenses/inter-LICENSE.txt 4.4 KB
- assets/review/licenses/tailwindcss-LICENSE.txt 1.0 KB
- assets/review/review.css 30 KB
- profiles/viverse.md 4.5 KB
- references/audit-and-quality.md 3.7 KB
- references/browser-adapters.md 7.9 KB
- references/data-schema.md 5.9 KB
- references/evidence-review.md 8.7 KB
- references/example-human-ai-review.md 6.2 KB
- references/human-demo.md 13 KB
- references/orchestration.md 2.8 KB
- references/play-protocol.md 12 KB
- references/report-layout.md 4.4 KB
- scripts/_common.py 832 B runs code
- scripts/contact_sheets.py 2.8 KB runs code
- scripts/frame_diff.py 1.7 KB runs code
- scripts/observed_kind_mismatches.py 2.0 KB runs code
- scripts/record_input.py 4.5 KB runs code
- scripts/review_sheet.py 13 KB runs code
- scripts/review-style/.gitignore 14 B
- scripts/review-style/copy-font.mjs 425 B runs code
- scripts/review-style/input.css 10 KB
- scripts/review-style/package-lock.json 34 KB
- scripts/review-style/package.json 515 B
- scripts/selftest.py 4.6 KB runs code
- scripts/spot_check.py 1.7 KB runs code
- scripts/test_review_sheet.py 7.8 KB runs code
- scripts/validate_descriptions.py 5.9 KB runs code
- templates/agent_brief.md 2.4 KB
- templates/plan.example.json 356 B
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.
- 20d ago First seen · 127 lines · 187 tokens per session scan A 32b0585d7acb
agent-playtest is a skill published in the GitHub repository jamesyang124/agent-skills (2 stars, last pushed 7d ago), licensed MIT. It adds 187 tokens to every session and 2,251 once invoked, about $0.0007 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 2026-10-06.
Other skills, from other repositories
debug-optimize-lcp
Guides debugging and optimizing Largest Contentful Paint (LCP) using Chrome DevTools MCP tools. Use this skill whenever the user asks about LCP performance, slow page loads, Core Web Vitals optimization, or wants to understand why their page's main content takes too long to appear. Also use when the user mentions…
interactive-login
Log a CLI in with a browser/device-code flow (aws sso, gh, glab, gcloud, …) and save the result to the keychain so later commands can request it through execute.credentials.
pinchtab-mcp
Use this skill when a task requires browser automation through PinchTab's MCP server connected to a remote browser instance. Covers navigation, element interaction, data extraction, form filling, multi-step flows, and session management via MCP tools.
azure-messaging-webpubsub-java
Build real-time web applications with Azure Web PubSub SDK for Java. Use when implementing WebSocket-based messaging, live updates, chat applications, or server-to-client push notifications.
cua-driver
Use Cua Driver for desktop or browser tasks that are awkward or unavailable through Bash/APIs, or when the user explicitly wants GUI interaction: app testing, visual bug reproduction, form filling, calendar entry, screenshots, and demo recording. Also covers Cua setup; not OpenAI Codex Computer Use or web research.
open-design-browser
Inspect and exercise local HTML design artifacts and library prototype previews with the existing agent-browser CLI, including screenshots, interaction checks, and rendered DOM snapshots. Use when rendered evidence helps author or review a design; excludes browser installation, remote services, and unrelated browser…