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/andreisilva1/binky/agents-mdgit clone --depth 1 https://github.com/andreisilva1/binkyWhat 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 | $0.00714 | $0.00714 |
| Opus 5 | $0.00357 | $0.00357 |
| Sonnet 5 | $0.00143 | $0.00143 |
| Haiku 4.5 | $0.00071 | $0.00071 |
Grade A, and why
binky AGENTS.md 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 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
2. Do the work; hit the service through its URL (or `binky curl` from a shell). How it starts
The opening of the file, as written. The whole thing — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Binky for agents
Binky gives each git worktree its own isolated local world — ports, env, a private
database slice, and stable URLs — so several agents can work on the same repo in
parallel without colliding on ports, databases, or .env. You call Binky; it never
drives you.
When to reach for it
- You're about to run a branch's dev server / migrations while another branch is already up.
- You need a throwaway but real database for a task, not a mock.
- You're one of several agents pointed at the same repository.
If you only ever have one world up at a time, you don't need Binky.
Install it as a tool
One line in Claude Code:
claude mcp add binky -- binky mcp
Or with nothing installed, via uvx:
claude mcp add binky -- uvx --from "binky[mcp]" binky mcp
Any MCP client takes the same stdio server as a config block:
{ "mcpServers": { "binky": { "command": "uvx", "args": ["--from", "binky[mcp]", "binky", "mcp"] } } }
Binky is also in the MCP Registry as
io.github.andreisilva1/binky, for clients that resolve servers from it.
The tools
binky mcp exposes the same verbs the CLI calls, against the same machine-global daemon:
| tool | signature | does |
|---|---|---|
up |
up(worktree, root="", group="") |
Bring a worktree's isolated world online (starts the daemon if needed). Returns its services, ports, and URLs. |
down |
down(worktree, root="") |
Stop a worktree's processes; keeps its files and ports. |
status |
status() |
List every worktree on this machine with its state, health, and ports. |
env |
env(worktree, root="") |
A worktree's live coordinates — service URLs and adapter addresses — to hand to a process or a test run. |
verify |
verify(worktree, root="") |
Run the project's agents.verify checks against a worktree; returns the pass/fail verdict. |
root defaults to the current working directory — pass it when you're operating on a
project elsewhere.
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 · 66 lines · 714 tokens per session scan A 3e971a33517d
binky AGENTS.md is an instructions file published in the GitHub repository andreisilva1/binky (1 stars, last pushed 24d ago), licensed Apache-2.0. It adds 714 tokens to every session, about $0.0036 per session on Opus 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.
Other instructions, from other repositories
mcp-context-forge AGENTS.md
AGENTS.md instructions for IBM/mcp-context-forge, covering agents.md, project overview, project structure, essential commands and setup.
pueue AGENTS.md
Instructions for Nukesor/pueue: This project does not approve of large-scale work done by LLMs.
api AGENTS.md
Instructions for vulnersCom/api, covering agents.md — using the vulners python sdk, install & authenticate, two ways to use it, 1. as a library and 2. as an mcp server.
yandex-tracker-mcp AGENTS.md
Instructions for aikts/yandex-tracker-mcp, covering agents.md, project overview, commands, architecture and talking to the tracker api.
RepoYeti AGENTS.md
Instructions for LunarWerxs/RepoYeti, covering agents.md, the one thesis, layout, before you push and daemon.
isrvd AGENTS.md
Instructions for rehiy/isrvd, covering agents.md — isrvd agent 操作指南, 1) 指令优先级, 2) 工作流, 3) 项目架构 and 实际依赖方向.