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/gszr/lunar/agents-mdgit clone --depth 1 https://github.com/gszr/lunarWrote 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/gszr/lunar/agents-md)<a href="https://agentmods.dev/instructions/gszr/lunar/agents-md"><img src="https://agentmods.dev/badge/instructions/gszr/lunar/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 | $0.00459 | $0.00459 |
| Opus 5 | $0.00230 | $0.00230 |
| Sonnet 5 | $0.00092 | $0.00092 |
| Haiku 4.5 | $0.00046 | $0.00046 |
Grade A, and why
lunar 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 5d 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.
What it actually says
How to work here
Lunar is a small Rust host with a Lua guest. Keep it that way.
Our goals
- We are a barebones coding harness.
- We aim at transparency context efficiency: no hidden system prompts that waste tokens and confuse the LLM.
Before you type
- Read
CONTEXT.md. Locked decisions stay locked. If something is unclear, ask. - Look at the code — and at Pi when the behavior is “like Pi” — before asking. If the tree answers it, do not ask.
- Unsettled product shape: ask one question, recommend an answer, wait.
- Once the shape is agreed, or the user says implement, stop asking and ship a small slice. Iterate later.
Code
- Simplest thing that works. No speculative abstraction, no “for later” types, no unused flags.
- Small modules, small surface. Split a file before it becomes a junk drawer.
mainstays a thin entry. - Match the code that is already here. Same names, same patterns, same density.
- Precise names from the domain.
/missionnot/session. Do not invent vocabulary. - Do not add workflow to the binary. Do not mix the Lua config path with the env path.
Scope
- Smallest useful diff for this request. One vertical slice you can run.
- Do not “improve” nearby code, comments, or formatting.
- Do not touch unrelated files. Leave workflow, generated, and personal files alone.
Check your work
- Investigate yourself. Do not ask the user to run something you can run.
- Prefer a real run over a mock. If it fails, find the root cause.
- Add a test only when it protects a behavior that could break. No tests that only prove a library works.
- When a product decision changes, update
CONTEXT.mdin the same turn. Keep the README short enough to start.
Git
- Commit only when asked.
- Conventional commits (
feat:,fix:,docs:,refactor:). - Separate commits for separate concerns.
- Run
cargo fmt --checkandcargo clippyand fix issues before committing.
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.
- 5d ago First seen · 45 lines · 459 tokens per session scan A 5e60cfd98f32
lunar AGENTS.md is an instructions file published in the GitHub repository gszr/lunar (10 stars, last pushed 4d ago), licensed MIT. It adds 459 tokens to every session, about $0.0023 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 2026-08-31.
Other instructions, from other repositories
teaql-agent-kit AGENTS.md
AGENTS.md instructions for teaql/teaql-agent-kit, covering agents.md — teaql agent kit and hard requirements.
concord-mcp CLAUDE.md
Claude Code instructions for Get-Concord-AI/concord-mcp, covering concord mcp — engineering guide, what this project is, coding rules (non-negotiable), the "never typecast" pattern for db rows and external json and architecture.
cezar AGENTS.md
Instructions for open-mercato/cezar, covering agents.md — working in this repository, zero config, changing a mechanism that already works, the http api and repository layout.
Trellis AGENTS.md
AGENTS.md instructions for mindfold-ai/Trellis, covering trellis instructions, gitnexus — code intelligence, always do, never do and resources.
Albatross AGENTS.md
Instructions for morganlinton/Albatross, covering agent instructions, general and releases.
x-code-cli AGENTS.md
AGENTS.md instructions for woai3c/x-code-cli, covering agents.md, rules, commands, delivery workflow and architecture.