DESIGN.md@benchmark/v2A

DESIGN.md@benchmark/v2 from pilotspace/add is a DESIGN.md file: the design rules coding agents follow in that repo.

Install it, pinned
mdr add pilotspace/add/benchmark/v2/DESIGN.md@git:20260710.cd78268

Pins git:20260710.cd78268 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 14 of 14 checks passed
Kind
DESIGN.md file
Repository
pilotspace/add
Path
benchmark/v2/DESIGN.md
Versions recorded
1
Latest version
git:20260710.cd78268, committed Jul 10, 2026
Size
5,642 bytes
Repository stars
19
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

# add-bench v2 — the trust-dimension benchmark (EXECUTED — WV1+WV2 rep0 complete, ledgers in `../results/`; WV3–WV6 remain gated)

## Why v2

v1's WM1 is a simple greenfield one-shot — the optimal case for a lightweight spec flow, and
spec-kit rightly wins it on cost (~3.3× cheaper at equal correctness, pinned-sonnet campaign,
`../results/2026-07-sonnet-campaign.md`). But v1 never measured the dimensions ADD's ceremony
exists to buy: **regression safety, gaming resistance, resumability, evolution under
contradiction, security floors, and traceability**. v2 measures those — fairly. Every arm
gets identical prompts, model pin, and budgets; every metric below is **deterministic where
possible** (the v1 LLM fidelity judge is retired as primary — see Meter fixes).

An honest outcome clause, up front: if spec-kit also holds these floors, v2 reports that.
The claim v2 tests is ADD's actual value proposition — *"trust the result without re-reading
the diff, across a codebase's lifetime"* — not a predetermined win.

## Meter fixes (prerequisites, carried from v1 findings)

1. **Deterministic fidelity oracle**: per-WM HTTP/CLI probe suites (endpoint × payload ×
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add pilotspace/add/benchmark/v2/DESIGN.md@sha256:78e3967648c65974

The line at the top of this page pins the label git:20260710.cd78268, 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_bexoczbdsf3pudkq.svg)](https://markdownregistry.com/a/art_bexoczbdsf3pudkq)

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

Versions

The registry has recorded 1 version of this file, committed on Jul 10, 2026.

versioncommittedcommitsizeaudit
git:20260710.cd78268 latest2026-07-10 cd78268 5,642 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 (5642 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

pilotspace/add · 19 stars · license MIT · pushed 2026-11-02 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_bexoczbdsf3pudkq
GET https://markdownregistry.com/api/v1/resolve?ref=pilotspace/add/benchmark/v2/DESIGN.md
GET https://markdownregistry.com/api/v1/blob/78e3967648c65974b886015d8725ed2cc3b7cfd3f3ba35c8d2170e9badcb1bcf

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: what DESIGN.md is and how agents use it · which agent file does what.

More from pilotspace/add

add SKILL.md skill
pilotspace/add · .claude/skills/add/SKILL.md · Use this skill when the user wants AI work they can trust without watching every step: the AI fixes direction first…
v4.1.0 · audit A · 19 stars
persona-author SKILL.md skill
pilotspace/add · .claude/skills/add/persona-author/SKILL.md · Author or improve an ADD-method persona file (a .add/personas/ slug.md) — the project-fit requirements LENS the ADD…
git:20260930.ae2c87d · audit A · 19 stars
AGENTS.md AGENTS.md file
pilotspace/add · AGENTS.md
git:20260930.ae2c87d · audit A · 19 stars
CLAUDE.md CLAUDE.md file
pilotspace/add · CLAUDE.md
git:20260930.ae2c87d · audit A · 19 stars
add SKILL.md skill
pilotspace/add · add-method/skill/add/SKILL.md · Use this skill when the user wants AI work they can trust without watching every step: the AI fixes direction first…
v4.1.0 · audit A · 19 stars
persona-author SKILL.md skill
pilotspace/add · add-method/skill/add/persona-author/SKILL.md · Author or improve an ADD-method persona file (a .add/personas/ slug.md) — the project-fit requirements LENS the ADD…
git:20260930.ae2c87d · audit A · 19 stars
add SKILL.md skill
pilotspace/add · add-method/src/add_method/_bundled/skill/add/SKILL.md · Use this skill when the user wants AI work they can trust without watching every step: the AI fixes direction first…
v4.1.0 · audit A · 19 stars
persona-author SKILL.md skill
pilotspace/add · add-method/src/add_method/_bundled/skill/add/persona-author/SKILL.md · Author or improve an ADD-method persona file (a .add/personas/ slug.md) — the project-fit requirements LENS the ADD…
git:20260930.ae2c87d · audit A · 19 stars
add SKILL.md skill
pilotspace/add · benchmark/arms/variants/add-4-audited/SKILL.md · Use this skill when the user wants AI work they can trust without watching every step: the AI fixes direction first…
v4.0.0 · audit A · 19 stars
add SKILL.md skill
pilotspace/add · benchmark/arms/variants/add-4-lean/SKILL.md · Use this skill when the user wants AI work they can trust without watching every step: the AI fixes direction first…
v4.0.0 · audit A · 19 stars
add SKILL.md skill
pilotspace/add · benchmark/arms/variants/add-4-probe/SKILL.md · Use this skill when the user wants AI work they can trust without watching every step: the AI fixes direction first…
v4.0.0 · audit A · 19 stars

Every file in pilotspace/add

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