Home / ccoalm / ccl-skills · GitHub

AGENTS.md@specs/042-skill-corpus-optimization/evidenceA

AGENTS.md@specs/042-skill-corpus-optimization/evidence from ccoalm/ccl-skills is an AGENTS.md file: the standing instructions coding agents read in that repo.

Install it, pinned

mdr add ccoalm/ccl-skills/specs/042-skill-corpus-optimization/evidence/AGENTS.md@git:20260824.5498662

Pins git:20260824.5498662 in mdr.lock. Pin the fingerprint instead.

Audit grade: A, 14 of 14 checks passed
Kind: AGENTS.md file
Repository: ccoalm/ccl-skills
Path: specs/042-skill-corpus-optimization/evidence/AGENTS.md
Versions recorded: 1
Latest version: git:20260824.5498662, committed Aug 24, 2026
Size: 3,832 bytes
Repository stars: 6
License: Apache-2.0

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

# evidence Agent Contract

042 轮的测量工件。本目录与前几轮不同的地方在于：**它记录了同一个问题被三种方法测出三个不同答案的全过程**，其中两个已作废。所以第一条规则不是"怎么跑"，是"哪些数不能引用"。

## 已作废的读数（引用前必看）

| 读数 | 出处 | 为什么作废 |
| --- | --- | --- |
| 「L106 承重，10/10 → 0/10」 | `deletion-L106.json` / `.txt` | **构造性同义反复**。marker 被 `deletion_arms.rb` 的隔离闸逼成「只有该行提到的词」，删掉该行模型自然不说那个词——测的是删词，不是变差。结论碰巧与后来一致，但不作证据 |
| 「L120/L113」的正则读数 | `deletion-L120-L113.json` | 同上 |
| 批 1 三处改动的正则读数（`0/10 → 8/10` 一类） | `conflict-arms-*.json` 的 `status` 字段 | **量级失真**。marker 写在看过新文本之后，继承其词汇；旧臂用别的说法说对同一件事被判 miss（实证：回答写「剥离」，正则写「剥掉」）。**这些文件的 `out` 字段仍然有效**——`rejudge.py` 正是重判它们 |
| 「零成本对调顺序有效」 | `order-probe.json` | 不复现（轮 1 discard 4→2，轮 2 持平），未采纳 |

**有效的读数**：`rejudge-verdicts.json`（按义务重判 180 条已存输出）、`frozen-rubric-run.json`（冻结判据 + 双向解离）、`final-verify.json`（对最终落地文本的无判据复验）。

## 规则

- **落地后冻结。** 不得为让后来的结论更好看而编辑、重判、裁剪或重生成这些文件。要推翻某个读数，就放新的测量并说明，像本目录已经做过三次的那样。
- **这些 runner 是证据工具，不是仓库闸。** 全部调用活模型、非确定性，绝不接入 `make test` 或任何阻断检查。这里跑绿不证明仓库任何性质。
- **`frozen-rubric.json` 不得事后编辑。** 它由独立通道在不知道任何删除候选的前提下产出，并在本目录任何一次冻结判据运行之前提交（`a82058c`）。判据一旦随结果调整，整套测量的意义就没了。其已知局限（O3 把义务与制品命名捆在一句）记在 `frozen-rubric.md`，**记而不改**。
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint

mdr add ccoalm/ccl-skills/specs/042-skill-corpus-optimization/evidence/AGENTS.md@sha256:b79d05d8d53e5f77

The line at the top of this page pins the label git:20260824.5498662, 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](https://markdownregistry.com/badge/art_l2et6qznfrkvgmm7.svg)](https://markdownregistry.com/a/art_l2et6qznfrkvgmm7)

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

Versions

The registry has recorded 1 version of this file, committed on Aug 24, 2026.

version	committed	commit	size	audit
git:20260824.5498662 latest	2026-08-24	5498662	3,832 B	A	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 (3832 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

ccoalm/ccl-skills · 6 stars · license Apache-2.0 · pushed 2026-11-05 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_l2et6qznfrkvgmm7
GET https://markdownregistry.com/api/v1/resolve?ref=ccoalm/ccl-skills/specs/042-skill-corpus-optimization/evidence/AGENTS.md
GET https://markdownregistry.com/api/v1/blob/b79d05d8d53e5f774ed1d08f0ebcb8e4fe606ec7f44ad82c72112ccc100c3721

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.