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 mattmre/EVOKORE-MCP-PUBLIC --skill browsergit clone --depth 1 https://github.com/mattmre/EVOKORE-MCP-PUBLICWrote 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/mattmre/evokore-mcp-public/browser)<a href="https://agentmods.dev/skills/mattmre/evokore-mcp-public/browser"><img src="https://agentmods.dev/badge/skills/mattmre/evokore-mcp-public/browser/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/mattmre/evokore-mcp-public/browser"><img src="https://agentmods.dev/badge/skills/mattmre/evokore-mcp-public/browser.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 | $0.00030 | $0.01681 |
| Opus 5.5 | $0.00012 | $0.00672 |
| Sonnet 5.5 | $0.00006 | $0.00336 |
| Haiku 4.5 | $0.00003 | $0.00168 |
Grade A, and why
browser scanned grade A with 1 finding 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 2mo 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
If the task is a single unauthenticated HTTP GET, prefer a plain `fetch`/`curl`-style tool instead of launching a browser. How it starts
The opening of the file, as written. The whole thing — 171 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Browser Automation Skill
AI-optimized browser automation using abbreviated element refs (93% context reduction vs verbose XPath/CSS selectors) and isolated multi-session contexts for parallel, side-effect-free flows.
Trigger
Use this skill when:
- Automating multi-step web UI flows (login, checkout, dashboard drilldowns)
- Scraping structured content from JS-rendered pages
- Running E2E UI tests against a deployed or local web app
- Orchestrating multiple browser tabs or parallel sessions (A/B flows, multi-account checks)
- Capturing screenshots, PDFs, or intercepting network requests as part of a verification pipeline
If the task is a single unauthenticated HTTP GET, prefer a plain fetch/curl-style tool instead of launching a browser.
Abbreviated Ref System
Verbose selectors dominate browser-automation context windows. Each document.querySelector('button[data-testid="submit-btn"]') burns tokens on every reference. Abbreviated refs replace these with short stable IDs assigned at snapshot time.
How It Works
At session or page snapshot, the browser tool enumerates interactive elements and returns a mapping table:
| Ref | Role | Label / Text | Selector (internal) |
|---|---|---|---|
#b1 |
button | "Submit" | button[data-testid="submit-btn"] |
#b2 |
button | "Cancel" | button.cancel-btn |
#i1 |
input | "Email" | input[name="email"] |
#i2 |
input | "Password" | input[name="password"] |
#a1 |
link | "Forgot password" | a[href="/reset"] |
The agent then operates in abbreviated-ref space:
click #b1
fill #i1 "[email protected]"
fill #i2 "hunter2"
click #b1
Under the hood, the tool resolves #b1 → real selector and executes. This trades ~100-byte selectors for 3-char refs — a 93% context reduction on typical flows.
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.
- 2mo ago First seen · 171 lines · 30 tokens per session scan A 96b4f686afce
browser is a skill published in the GitHub repository mattmre/EVOKORE-MCP-PUBLIC (3 stars, last pushed 5mo ago), licensed MIT. It adds 30 tokens to every session and 1,681 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
browser compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| browsermattmre/EVOKORE-MCP-PUBLIC | Skill | 3 | 5mo ago | MIT | A | 30 | Claude Code, Codex |
| webapp-testingLiHongwei-cn/lihongwei-cn | Skill | 5 | 1mo ago | MIT | A | 17 | Claude Code, Codex |
| webapp-testingphuonghx/aim-cli | Skill | 1 | 4mo ago | MIT | A | 74 | Claude Code, Codex |
| browser-automationLiHongwei-cn/lihongwei-cn | Skill | 5 | 1mo ago | MIT | A | 42 | Claude Code, Codex |
| e2e-testing-patternswshobson/agents | Skill | 42k | 4d ago | MIT | A | 51 | Claude Code |
| browser-testing-with-screenshotsAgentWorkforce/relay | Skill | 1.1k | today | Apache-2.0 | A | 32 | Claude Code |
| browser-qa-deliveryjuspay/xyne-spaces | Skill | 826 | today | Apache-2.0 | A | 30 | Claude Code, Codex |