Borrowing it
Nothing to install: this file belongs to lna-lab/distill-kura. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/lna-lab/distill-kura/master/AGENTS.mdgit clone --depth 1 https://github.com/lna-lab/distill-kuraWrote 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/lna-lab/distill-kura/agents-md)<a href="https://agentmods.dev/instructions/lna-lab/distill-kura/agents-md"><img src="https://agentmods.dev/badge/instructions/lna-lab/distill-kura/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/lna-lab/distill-kura/agents-md"><img src="https://agentmods.dev/badge/instructions/lna-lab/distill-kura/agents-md.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.03021 | $0.03021 |
| Opus 5 | $0.01510 | $0.01510 |
| Sonnet 5 | $0.00604 | $0.00604 |
| Haiku 4.5 | $0.00302 | $0.00302 |
Grade A, and why
distill-kura 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 6d 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.
Copies of this mod
1 near-identical copy found in the catalogue:
- distill-kura AGENTS.md — 100% identical, 13 lines differ
How it starts
The opening of the file, as written. The whole thing — 232 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — working on distill-kura
Instructions for an agent (or a person) changing this codebase. Hosts that read
AGENTS.md — DeepSeek Harness via dsh-agent-instructions, Claude Code, others — pick
this up automatically.
What this project is
A long-term memory for agents: recall by meaning, writing gated by evidence, several independent stores so an agent mode change is a memory change. Standard library only, Python ≥ 3.11. No dependencies is a feature — it lets the whole thing be dropped next to any host, and it keeps the trust surface small for something that decides what an agent believes.
Boundaries
docs/TRUST.md states what a store boundary is and is not, and the honesty of that
statement is load-bearing. Two rules follow from it and must not be weakened:
Every lookup resolves INTO slug_set(). Never build a path from a caller-supplied
name and check whether it exists — that was the hole: GET /memory/..%2Fprivate%2Fsecret
returned another store's memory. Containment is membership in a set, not a blocklist of
characters. contained() (realpath + commonpath) sits behind it as defence in depth.
Explicit reads are exact; only a MODEL's pick is fuzzy. read_exact() for a slug from
a person, a kura_read call, or an HTTP route. Fuzzy resolve() stays for thinker picks
and [[links]], where every candidate comes from the slug set — a deliberate deviation
from "links exact only", because in-store fuzzy resolution demonstrably connects real
links ([[brain-memory]] → _study/brain-memory) and cannot leave the store.
Every path that writes into a store asks the policy. Not just remember_direct and
pour_verified — an adversarial pass found tidy(), Loom.persist() and init_files()
writing into a frozen store, and kura weave --no-model destroying a memory's body via
cloth_path. When you add a code path that touches a store directory, the question to
answer in review is not "is this a memory?" but "would this run on a frozen store?".
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.
- 6d ago Changed · +7 lines · +146 tokens per session bf20f636c32d
- 10d ago First seen · 225 lines · 2,875 tokens per session scan A 42d4b23bd5d9
distill-kura AGENTS.md is an instructions file published in the GitHub repository lna-lab/distill-kura (48 stars, last pushed 4d ago), licensed MIT. It adds 3,021 tokens to every session, about $0.0151 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
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.