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/aleksandr-voitenko/agentic-lore-coding/agents-mdgit clone --depth 1 https://github.com/aleksandr-voitenko/agentic-lore-codingWhat 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 | $0.08872 | $0.08872 |
| Opus 5 | $0.04436 | $0.04436 |
| Sonnet 5 | $0.01774 | $0.01774 |
| Haiku 4.5 | $0.00887 | $0.00887 |
Grade A, and why
agentic-lore-coding 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 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.
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 — 854 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Introduction
You are working in an Agentic Lore Coding environment.
Instruction strength
Use these meanings throughout this file:
- Must means required unless impossible in the current environment or explicitly overridden by the user.
- Should means expected for normal work, but may be skipped when repository context makes it inappropriate.
- Prefer means a default style or design preference, not a hard rule.
When instructions conflict, follow safety and repository-specific requirements first, then task traceability, then general preferences.
Software development principles
Core guidelines
Do not make unsupported guesses.
Start non-trivial work with research or planning. Turn vague input into a concrete implementation plan before writing code. Identify important working assumptions and separate them from open questions that would change the plan.
Consider alternatives when a decision affects architecture, behavior, public APIs, data models, dependencies, performance, security, accessibility, deployment, tooling, or maintainability.
Keep changes atomic. Do not mix unrelated behavior changes, refactors, formatting, dependency updates, generated-file updates, or UI cleanup in one task. Minimize the footprint of a change and use existing abstractions, owners, or integration points when they fit cleanly.
If a separate unrelated issue is discovered, leave it unchanged and notify the user. Include it in the current task only when necessary to complete or verify the requested change safely.
Do not bypass, remove, or alter environment checks, sandbox checks, feature gates, test guards, CI guards, or safety-related conditions unless the task explicitly requires it and the reason is understood and documented.
Semantic sanity gate
Before implementing any task, check whether the literal request makes sense for the current product, game, domain, and user-facing experience.
Do not convert a nonsensical request into a plausible feature by inventing an interpretation. The existence of a low-risk implementation does not make the assumption safe.
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 · 854 lines · 8,872 tokens per session scan A b7fe6a284bcc
agentic-lore-coding AGENTS.md is an instructions file published in the GitHub repository aleksandr-voitenko/agentic-lore-coding (5 stars, last pushed 2mo ago), licensed MIT. It adds 8,872 tokens to every session, about $0.0444 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
har AGENTS.md
Instructions for os-factory/har, covering har — agent development guide, har / agent environment, harnesses in this repo, harness workflow (dogfooding) and run history.
sprawling AGENTS.md
Instructions for 2youg1/sprawling, covering agents.md — how work is done in this repository, the loop, read before you write, one change, five steps and the rules a machine holds.
aster AGENTS.md
Instructions for Zfinix/aster, covering aster, global rules, user-facing copy, rust conventions and module size.
cctop AGENTS.md
Instructions for st0012/cctop, covering agents.md - development guide for cctop, scope and sources, required development rules, driver workflow and github and pr rules.
sprawling CLAUDE.md
Instructions for 2youg1/sprawling: See AGENTS.md. It is the whole instruction set for this repository — the loop, what to read first, the five steps of one change, the rules a machine holds, and the language and commit conventions.
cctop CLAUDE.md
Instructions for st0012/cctop: This repository keeps shared agent instructions in AGENTS.md. Read AGENTS.md now and follow it as the source of truth for cctop development.