AGENTS.md@src/coreA

AGENTS.md@src/core from evanzhang008/open-walnut is an AGENTS.md file: the standing instructions coding agents read in that repo.

Install it, pinned
mdr add evanzhang008/open-walnut/src/core/AGENTS.md@git:20261004.e7e6481

Pins git:20261004.e7e6481 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 14 of 14 checks passed
Kind
AGENTS.md file
Repository
evanzhang008/open-walnut
Path
src/core/AGENTS.md
Versions recorded
5
Latest version
git:20261004.e7e6481, committed Oct 4, 2026
First version recorded
Sep 2, 2026
Size
4,838 bytes
Repository stars
34
License
MIT

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# Core Layer — Quick Reference

**Full implementation details: `.claude/skills/walnut-core-internals/SKILL.md`** (task model,
phase lifecycle, session monitoring, message queue, JSONL patterns, compaction). Load that skill
before non-trivial work in `src/core/`.

**Decision records live in [`docs/decision/`](../../docs/decision/).** Read the relevant record
before rebuilding or "fixing" anything it says was deliberately removed. In particular, read
the [self-report](../../docs/decision/summarizer-self-report.md) and
[session-end](../../docs/decision/no-session-end-gist.md) decisions before changing
turn-complete task updates or session lifecycle hooks.

**Search/indexing:** the engine is `src/lib/hybrid-search/` (self-contained library: tokenizer,
FTS5 schema, two keyword lanes, vector rescore, embed worker) with `src/core/search/` as the
walnut adapter (serializers, event-bus wiring, backfill). Read
[`src/lib/hybrid-search/README.md`](../lib/hybrid-search/README.md) before changing interactive
search, the tokenizer, embedding models, or the indexing path. The library must keep ZERO walnut
imports (`tests/lib/hybrid-search-boundary.test.ts` enforces it).
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add evanzhang008/open-walnut/src/core/AGENTS.md@sha256:738d4d58d1d2998c

The line at the top of this page pins the label git:20261004.e7e6481, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_totsw2rn64sf4x4d.svg)](https://markdownregistry.com/a/art_totsw2rn64sf4x4d)

3 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 5 versions of this file, between Sep 2, 2026 and Oct 4, 2026, about one every 8 days.

versioncommittedcommitsizeaudit
git:20261004.e7e6481 latest2026-10-04 e7e6481 4,838 BA view · diff
git:20260929.89b1b352026-09-29 89b1b35 4,209 BA view · diff
git:20260922.954bb922026-09-22 954bb92 3,161 BA view · diff
git:20260912.1cb823b2026-09-12 1cb823b 2,652 BA view · diff
git:20260902.792e4d82026-09-02 792e4d8 2,384 BA view

Audit of the latest version

A  14 of 14 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Size between 200 bytes and 200 KB (4838 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

evanzhang008/open-walnut · 34 stars · license MIT · pushed 2026-10-22 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_totsw2rn64sf4x4d
GET https://markdownregistry.com/api/v1/resolve?ref=evanzhang008/open-walnut/src/core/AGENTS.md
GET https://markdownregistry.com/api/v1/blob/738d4d58d1d2998c03b9d797e2cf80812a65c83ad2186663bd3019182a07a28d

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: how to write a good AGENTS.md · AGENTS.md vs CLAUDE.md · which agent file does what.

More from evanzhang008/open-walnut

test-and-verify-walnut SKILL.md skill
evanzhang008/open-walnut · .claude/skills/test-and-verify-walnut/SKILL.md · 4-agent pipeline: design + implement code tests + Playwright UI tests + quality gate. Reads walnut-console-ops first.
git:20260923.a2cf102 · audit A · 34 stars
walnut-console-ops SKILL.md skill
evanzhang008/open-walnut · .claude/skills/walnut-console-ops/SKILL.md · Operating the Walnut web UI (localhost:3456) via Playwright MCP tools — understanding the UI layout, chatting with main…
git:20260923.a2cf102 · audit A · 34 stars
walnut-core-internals SKILL.md skill
evanzhang008/open-walnut · .claude/skills/walnut-core-internals/SKILL.md · Deep implementation details for src/core/ — task model (5 text fields, phase lifecycle, dependencies, session slots…
git:20260929.89b1b35 · audit A · 34 stars
walnut-markdown-rendering SKILL.md skill
evanzhang008/open-walnut · .claude/skills/walnut-markdown-rendering/SKILL.md · How Walnut turns model text into rendered output — the two marked instances and the CJK retunes registered on both, why…
git:20261013.45aeb25 · audit A · 34 stars
walnut-testing SKILL.md skill
evanzhang008/open-walnut · .claude/skills/walnut-testing/SKILL.md · Testing details for the Walnut repo — 5-tier pyramid (unit/integration/e2e/browser/live), per-tier vitest configs &…
git:20260923.a2cf102 · audit A · 34 stars
walnut-web-frontend SKILL.md skill
evanzhang008/open-walnut · .claude/skills/walnut-web-frontend/SKILL.md · Implementation details for web/src/ (React SPA) — session chat single-timeline model (streaming blocks, render-filter…
git:20261004.76c60f1 · audit A · 34 stars
AGENTS.md AGENTS.md file
evanzhang008/open-walnut · AGENTS.md
git:20261020.83cc878 · audit A · 34 stars
CLAUDE.md CLAUDE.md file
evanzhang008/open-walnut · CLAUDE.md
git:20260706.0a2d7f0 · audit B · 34 stars
hello-greetings SKILL.md skill
evanzhang008/open-walnut · examples/plugins/hello-walnut/skills/hello-greetings/SKILL.md · Greet someone using the Hello Walnut plugin's tool. Use when the user asks for a greeting demo, says "greet X", or…
git:20260821.22b34fe · audit A · 34 stars
walnut-demo SKILL.md skill
evanzhang008/open-walnut · examples/plugins/walnut-demo/skills/walnut-demo/SKILL.md · Read the Walnut Plugin Demo's state and explain which Plugin capability does what.
git:20260824.e042aac · audit A · 34 stars
walnut-rhythm SKILL.md skill
evanzhang008/open-walnut · plugin-store/walnut-rhythm/skills/walnut-rhythm/SKILL.md · Start, stop and read Rhythm focus blocks (pomodoros tied to a task) and answer or snooze its stand-up reminder. Use…
git:20260925.ebb4db2 · audit A · 34 stars
restore-backup SKILL.md skill
evanzhang008/open-walnut · skills/restore-backup/SKILL.md · Hands-on disaster-recovery agent for Open Walnut S3 backups. Paste into your OWN Claude Code session — it finds your…
git:20260815.05be0cd · audit A · 34 stars

Every file in evanzhang008/open-walnut

Browse by kind, by grade A, or by owner.