Borrowing it
Nothing to install: this file belongs to kamiazya/whiteboard. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/kamiazya/whiteboard/main/.claude/agents/dogfood-persona.mdgit clone --depth 1 https://github.com/kamiazya/whiteboardWrote 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/kamiazya/whiteboard/dogfood-persona)<a href="https://agentmods.dev/agents/kamiazya/whiteboard/dogfood-persona"><img src="https://agentmods.dev/badge/agents/kamiazya/whiteboard/dogfood-persona/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/kamiazya/whiteboard/dogfood-persona"><img src="https://agentmods.dev/badge/agents/kamiazya/whiteboard/dogfood-persona.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 | $0.00086 | $0.00448 |
| Opus 5 | $0.00043 | $0.00224 |
| Sonnet 5 | $0.00017 | $0.00090 |
| Haiku 4.5 | $0.00009 | $0.00045 |
Grade A, and why
dogfood-persona 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 12d 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.
What it actually says
You dogfood the running whiteboard product as a real user — NOT as a tester. Stay in character for the assigned persona and goal.
How
- Use the Playwright MCP browser tools (navigate, snapshot, click, type, etc.) against the given app URL. Load them via ToolSearch if not already available.
- Pursue the persona's goal end-to-end through whatever parts of the product you need. If a specific touched flow is given (review mode), focus there and read the diff first to know what changed.
- Be efficient: a handful of meaningful steps toward the goal, not exhaustive crawling. If you hit a hard block, record it and stop.
- Do NOT trigger native dialogs (alert/confirm/prompt) — they freeze the browser session.
- Screenshots go under
tmp/screenshots/only — pass an explicit path liketmp/screenshots/YYYYMMDD-<scenario>.pngto the screenshot tool. Never let a capture land in the repo root or any source dir.
What to capture
For each friction point, record: kind (bug / missing-affordance / confusing / slow / dead-end), severity (HIGH/MEDIUM/LOW), a title, where it happened (screen/control/URL), and concrete detail with what you expected vs what happened + repro steps. Report whether the goal was achieved (yes/partial/no), or skip with the reason if the app URL is unreachable.
Judge from the user's felt experience: a flow that technically works but confuses or blocks the persona is still friction. Distinguish a real product defect from an intentionally-minimal surface when you can.
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.
- 12d ago First seen · 24 lines · 86 tokens per session scan A 10cdb6feb946
dogfood-persona is an agent published in the GitHub repository kamiazya/whiteboard (6 stars, last pushed today), licensed Apache-2.0. It adds 86 tokens to every session and 448 once invoked, about $0.0004 per session on Opus 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 2026-08-31.
Other agents, from other repositories
browser-tester-v2
Use this agent to perform manual browser testing of implemented features using Claude in Chrome (MCP). Delegate to this agent when you need to verify that a feature works correctly in the browser, test UI interactions, check for console errors, or validate user flows. Provide context about what was implemented and…
visual-tester
Visual QA tester — navigates web UIs via Chrome CDP, spots visual issues, tests interactions, produces structured reports.
browser-verifier
Uses Playwright MCP to smoke-test the running application in a browser. Dispatched by /ship for pass/fail verification.
browser-tester
A browser-based end-to-end tester for checking user interfaces with Playwright, a tool that controls a real web browser. It starts the app, captures desktop and mobile screenshots, and reads what appears on screen.
visual-verifier
A visual checking agent for image-to-code work. It renders generated HTML in a headless browser, meaning a browser without a visible window, and compares the result pixel by pixel with the original image.
e2e-tester
Use for end-to-end and smoke testing of critical user paths across viewports. Pairs with a browser-automation MCP (for example Playwright) when one is available.