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/spuentesp/lain/agents-mdgit clone --depth 1 https://github.com/spuentesp/lainWrote 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/spuentesp/lain/agents-md)<a href="https://agentmods.dev/instructions/spuentesp/lain/agents-md"><img src="https://agentmods.dev/badge/instructions/spuentesp/lain/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 | $0.00689 | $0.00689 |
| Opus 5 | $0.00345 | $0.00345 |
| Sonnet 5 | $0.00138 | $0.00138 |
| Haiku 4.5 | $0.00069 | $0.00069 |
Grade C, and why
lain AGENTS.md scanned grade C with 2 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 today.
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.
Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
site, so `curl … | bash` and direct invocation both work. Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
site, so `curl … | bash` and direct invocation both work. How it starts
The opening of the file, as written. The whole thing — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — notes for AI agents working in this Lain checkout
Background
The upstream v0.7.0 release tarball shipped a lain binary whose
--version reported lain 0.6.1. Root cause: Cargo.toml's
version field was never bumped from 0.6.1 when the v0.7.0
git tag was cut, so any binary built from that tree carried the
old version string. Functionally the binary behaved as v0.7.0;
only the version string was wrong.
Status (2026-09-02 → 2026-09-03)
- Local fix (commits
93d6344+ae2a527): bumpedCargo.tomlto0.7.0onfix/v0.7.0-version-bump, merged tomain, and rebuilt~/.local/lain/lainso--versioncorrectly reportslain 0.7.0. Original tarball binary is kept as~/.local/lain/lain.bak.0.6.1. - Upstream fix (tag
v0.7.1): bumpedCargo.tomlto0.7.1and pushed the tag so the release workflow publishes corrected binaries to GitHub Releases.server.json's top-level + nestedversionfields get updated automatically by the release workflow;Formula/lain.rbandnpm-shim/package.jsonare out of scope for this fix. - Installer fix (tag
v0.7.2):install.shhad a function-call-ordering bug — it invokedapply_noninteractive_defaultsat the top of the file before defining the function further down. Withset -e, that killed the script withcommand not foundbefore any work happened, which is why this environment always installed Lain manually.apply_noninteractive_defaultsis now defined above its call site, socurl … | bashand direct invocation both work.
After the v0.7.1 workflow completed, the official tarballs at
https://github.com/spuentesp/lain/releases/tag/v0.7.1 report
the correct version string. After the v0.7.2 workflow
completes, fresh installs no longer hit the silent-exit bug.
Re-installing the official tarball
install.sh from upstream will now install lain 0.7.2 (or
newer). Fresh installs work end-to-end via curl … | bash or
direct invocation — no more manual install dance.
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.
- today First seen · 51 lines · 689 tokens per session scan C 723573fdf9de
lain AGENTS.md is an instructions file published in the GitHub repository spuentesp/lain (6 stars, last pushed yesterday), licensed MIT. It adds 689 tokens to every session, about $0.0034 per session on Opus 5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-04.
Other instructions, from other repositories
ai-sdk CLAUDE.md
Claude Code instructions for open-metadata/ai-sdk, covering claude.md - metadata ai sdk, repository structure, quick commands, linting & formatting and testing.
harvest-mcp CLAUDE.md
Claude Code instructions for standardbeagle/harvest-mcp, covering claude.md, project overview, essential commands, development workflow and install dependencies (may fail in claude code environment).
openings-mcp AGENTS.md
Instructions for amikai/openings-mcp, a project described as: 💼 One MCP server to search job boards and company career sites.
brapi-mcp-server AGENTS.md
Instructions for cyanheads/brapi-mcp-server, covering agent protocol, what's next?, core rules, patterns and tool — connection bootstrap.
mcp-redmine-rest CLAUDE.md
Claude Code instructions for alsimoes/mcp-redmine-rest, covering claude.md, what this is, layout (src-layout), tool registration pattern and deliberately missing tools.
onplana-mcp-server GEMINI.md
Instructions for Onplana/onplana-mcp-server, covering onplana, gemini context, what onplana is, authenticating, hints when working with onplana data and common workflows.