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 instructions/plasmate-labs/plasmate/agents-mdgit clone --depth 1 https://github.com/plasmate-labs/plasmateWrote 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/instructions/plasmate-labs/plasmate/agents-md)<a href="https://agentmods.dev/instructions/plasmate-labs/plasmate/agents-md"><img src="https://agentmods.dev/badge/instructions/plasmate-labs/plasmate/agents-md.svg" alt="Measured on agentmods" 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 | $0.01534 | $0.01534 |
| Opus 5 | $0.00767 | $0.00767 |
| Sonnet 5 | $0.00307 | $0.00307 |
| Haiku 4.5 | $0.00153 | $0.00153 |
Grade A, and why
plasmate AGENTS.md 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 3d 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 — 157 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Plasmate Codebase Guide
This file is for AI coding agents (Cursor, Devin, Claude Code, Copilot, etc.). It tells you what the codebase does, how it is structured, and how to make changes safely.
What Plasmate is
Plasmate is a headless browser engine that compiles web pages into a Semantic Object Model (SOM) — structured JSON optimised for LLM consumption — instead of returning raw HTML. Output size is page-, configuration-, corpus-, and serialization-dependent. Retained v0.5.1 public-web snapshots measured median serialized-byte ratios of 9.98x over 83 successful non-JavaScript inputs and 9.32x over 82 successful JavaScript inputs, each from 98 attempted URLs. Those historical observations are not universal token, cost, latency, or task success claims. No API key, no cloud.
It runs as a CLI, a persistent daemon, an MCP server, and a CDP server.
Build
~/.cargo/bin/cargo build # debug
~/.cargo/bin/cargo build --release # release
Requires Rust 1.88+ (including --all-features). No system dependencies beyond a C linker.
Test
~/.cargo/bin/cargo test # all tests
~/.cargo/bin/cargo test som:: # SOM tests only
~/.cargo/bin/cargo test mcp:: # MCP tests only
RUST_LOG=debug ~/.cargo/bin/cargo test -- --nocapture # with logging
Run the suite rather than citing a hard-coded test count; the count changes as the repository evolves. All tests must pass before a PR.
Key directories
src/
main.rs CLI entry point (fetch, compile, diff, mcp, serve, daemon, screenshot)
mcp/
mod.rs MCP server, JSON-RPC router, session manager
tools.rs ALL MCP tool definitions + handlers (add new tools here)
sessions.rs Persistent browser session state
som/
mod.rs SOM data structures and serialisation
filter.rs apply_selector() — shared between CLI and MCP
compiler.rs HTML → SOM compiler (the core algorithm)
js/
runtime.rs V8-backed JS execution
pipeline.rs Full fetch+JS+compile pipeline
network/
fetch.rs HTTP client (reqwest)
sdk/python/ Python SDK (MCP client)
sdk/node/ Node.js SDK (MCP client)
integrations/ LangChain, LlamaIndex, Browser Use, etc.
packages/ som-parser-python, som-parser-node
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.
- 3d ago First seen · 157 lines · 1,534 tokens per session scan A 5977623fbd4e
plasmate AGENTS.md is an instructions file published in the GitHub repository plasmate-labs/plasmate (36 stars, last pushed 3d ago), licensed Apache-2.0. It adds 1,534 tokens to every session, about $0.0077 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 instructions, from other repositories
obscura AGENTS.md
Instructions for h4ckf0r0day/obscura, covering agents.md, build, rendering and stealth, no rendering, with rustls or stealth and test.
browser-harness AGENTS.md
Instructions for browser-use/browser-harness, covering code priorities and contributing.
browser-harness CLAUDE.md
Instructions for browser-use/browser-harness, a project described as: Browser Harness | Self-healing harness that enables LLMs to complete any task.
browser-debugger-cli CLAUDE.md
Instructions for szymdzum/browser-debugger-cli, covering claude.md, agent-friendly discovery (start here), essential patterns, commandrunner (src/commands/shared/commandrunner.ts) and error handling.
unbrowser CLAUDE.md
Instructions for protostatis/unbrowser, covering unbrowser, why this exists, architecture (locked), stealth mode (native, on by default) and host functions exposed to js.
starling AGENTS.md
Instructions for starling-browser/starling, covering agents — rules of engagement, tl;dr, 1. orient, 2. claim something unblocked and 3. read the package file and start working on main.