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 agentmods add skills/debugbase/glance/glance-testnpx skills add DebugBase/glance --skill glance-testgit clone --depth 1 https://github.com/DebugBase/glanceWhat 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 | $0.00080 | $0.00883 |
| Opus 5 | $0.00040 | $0.00441 |
| Sonnet 5 | $0.00016 | $0.00177 |
| Haiku 4.5 | $0.00008 | $0.00088 |
Grade A, and why
glance-test 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 2d 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 — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Glance E2E Browser Test
You run end-to-end browser tests using Glance MCP tools. You have a real Chromium browser at your disposal.
Prerequisites
Glance MCP must be configured. If mcp__browser__browser_navigate is not available, tell the user:
claude mcp add glance -- npx glance-mcp
Workflow
1. Get the target
Ask for the URL if not provided. Accept:
- Full URL:
https://example.com - Local:
localhost:3000 - Relative paths (prepend the known base URL)
2. Start session and navigate
mcp__browser__session_start — name: "e2e-{domain}"
mcp__browser__browser_navigate — url
3. Initial assessment
mcp__browser__browser_screenshot — see the page
mcp__browser__browser_snapshot — get DOM structure
mcp__browser__browser_console_messages — check for JS errors
4. Smart page discovery
From the snapshot, identify:
- Navigation links (sidebar, header, footer)
- Forms (login, register, contact, search)
- CTAs (buttons, links)
- Interactive elements (dropdowns, modals, tabs)
5. Test each page
For every discoverable page, run:
navigate → screenshot → assert key elements → check console → check network
Use test_scenario_run for multi-step flows:
{
"name": "Page: /login",
"steps": [
{"name": "Navigate", "action": "navigate", "url": "URL"},
{"name": "Page loaded", "action": "assert", "type": "exists", "selector": "h1"},
{"name": "Screenshot", "action": "screenshot", "screenshotName": "page-name"},
{"name": "No console errors", "action": "assert", "type": "consoleNoErrors"}
]
}
6. Test forms and auth
If login/register forms exist:
- Test with invalid data (expect error message)
- Test with valid data if credentials provided
- Verify redirects and session persistence
7. Generate report
Output a markdown table:
| Page | Steps | Pass | Fail | Issues |
|------|-------|------|------|--------|
| / | 5 | 5 | 0 | None |
| /login | 8 | 7 | 1 | Console error: ... |
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.
- 2d ago First seen · 126 lines · 80 tokens per session scan A 58127dc2f723
glance-test is a skill published in the GitHub repository DebugBase/glance (151 stars, last pushed 4mo ago), licensed MIT. It adds 80 tokens to every session and 883 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 2026-08-30.
Other skills, from other repositories
compiler
Use when converting a refined markdown E2E test case into a standard Playwright .spec.ts file. Triggers on "compile this test", "convert to playwright", or "generate spec from markdown". Requires a stable markdown test case and a running dev server.
author
Use when writing a new E2E browser test from a vague user description or feature requirement. Turns exploratory browser sessions into structured, reusable test cases.
refiner
Use when an E2E test case exists but needs stabilization through repeated execution. Runs the test multiple times, identifies flaky steps, and updates the test case and conventions until deterministic.
runner
Use when executing a stable E2E test case that has already been refined. Follows test steps exactly without deviation, reports pass/fail with evidence.
CBrowser
Cognitive Browser — AI browser automation that simulates real user cognition. 122 MCP tools, 21 personas, 25 cognitive traits, constitutional safety, cognitive journeys with LLM-free trace replay, screen capture to GIF/WebP/WebM with an AI-readable frame manifest (v18.69.3). USE WHEN browser automation OR cognitive…
playwright-cli
Automate browser interactions, test web pages and work with Playwright tests.