skills: Instructions file for Claude Code

AGENTS.md

skills AGENTS.md is an instructions file for Claude Code, Codex, OpenCode from wildcat-finance/skills. It costs 4,777 tokens per session, scanned A, original, Apache-2.0.

A repository instruction file for a collection of agent skills. It defines how agents select skills, follow shared rules, and handle evidence and work transitions.

In plain words
What is it for?
Use it when working in that skills repository, especially when selecting a skill, checking its required contract, or managing dependent work.
Why use it?
It prevents agents from treating every present file as active and sets a common process for safe, evidence-based work.

Instructions file for Claude CodeCodexOpenCode

Written for Claude Code and Codex and OpenCode: Claude Code plugin machinery, but also the file is AGENTS.md. Also seen: mentions AGENTS.md; mentions Codex.

This is wildcat-finance/skills's own configuration. It tells Claude Code, Codex and OpenCode how to work on skills itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything skills configures →

Reuse

Borrowing it

Nothing to install: this file belongs to wildcat-finance/skills. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/wildcat-finance/skills/main/AGENTS.md
Clone the repo
git clone --depth 1 https://github.com/wildcat-finance/skills

Made for: Claude Code, Codex, OpenCode.

Wrote 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.

agentmods badge for skills AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/wildcat-finance/skills/agents-md.svg)](https://agentmods.dev/instructions/wildcat-finance/skills/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/wildcat-finance/skills/agents-md"><img src="https://agentmods.dev/badge/instructions/wildcat-finance/skills/agents-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 4,777 This file is loaded in full into every session.
When invoked 4,777 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.04777 $0.04777
Opus 5 $0.02388 $0.02388
Sonnet 5 $0.00955 $0.00955
Haiku 4.5 $0.00478 $0.00478

Measured today against content hash 4a4d2db22e4b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

skills 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 today.

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.

AGENTS.md · 391 lines

How it starts

The opening of the file, as written. The whole thing — 391 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Instructions for local agents

This repository is the source distribution for Wildcat Labs Skills, the Shoggoth. A person can begin with README.md; an agent begins here.

The safe loading path is short:

  1. read the collective identity and shared Promise Machine contract;
  2. route the request through .agents/skills/promise-machine/SKILL.md;
  3. read the selected plugin's AGENTS.md; and
  4. read the one canonical SKILL.md that contract selects.

Files present in context are not automatically active skills. A router, manifest, worker brief, or generated installation copy also does not create a general capability or permission.

Collective identity

Before interpreting any reference to Shoggoth, including a shortened, altered, or affectionate form of the name, or resolving a user's collective form of address from available GitHub credentials, read the shared Shoggoth collective identity. That language affects how people address the agents and skills. It does not activate a skill, grant authority, or weaken the Promise Machine and target-repository instructions.

Promise Machine contract

Before selecting or running a skill, read the suite-wide Promise Machine contract. Its identity is promise-machine/v1. Each result authorises only the transition declared by its canonical skill; missing, stale or insufficient evidence blocks that dependent transition while leaving inspection, repair, rerun and safe exit available.

The root law is authored once. Plugin-local PROMISE_MACHINE.md files are generated installation copies and must remain byte-identical to it.

Marketplace boundaries

The seventeen plugins form one marketplace, not seventeen competing descriptions of the same job. Alexandria preserves lending inputs; Tabularium interprets preserved venue records; Probitas assembles a counterparty dossier. Lazarus preserves the finite historical Ethereum state and exact RPC traffic a test needs, while Ariadne binds a released artefact digest to its evidence. Berean holds a protocol agent's recorded answers to pinned corpora and preserved chain reads; it neither chunks documents nor preserves chain state itself. Pandects supplies reviewed credit laws, Hermes measures a single gas-optimisation class named by a rule from its pinned corpus, Hexaemeron controls a receipted delivery loop and holds each of its phases to a named skill, while Lemma stops after producing source-linked chunks. Homologia compares one pinned contract computation with one pinned off-chain mirror over declared vectors and preserves each divergence as a specimen, where Pandects supplies the economic laws and Lazarus the proved chain-side answers such a comparison consumes as evidence. Horos decides what an agent does not read. Janus checks what a contract hook may observe and change around a host action, where Pandects supplies the economic laws such a transition must preserve. Synkrisis owns the cross-run comparison boundary: it builds one checked cohort from declared run observations and infers only bounded relations between named events, and any operation whose runbook step has not landed refuses by name rather than guessing. All four current operations ship: cohort, diagnosis, rendering, and whole-path verification. Its findings stay bounded inferences: capture, redaction, receipt binding, causal triage, and every decision to act stay with their owners. Anamnesis keeps custody of audit findings and the changes that answered them: it admits a source only against an explicit rights basis, preserves the producer's bytes unchanged, and hands Elenchus a historical analogue and Synkrisis a checked cohort projection. The analogue is a hypothesis, so Elenchus still reproduces the present failure and still earns its own guard, and Synkrisis takes no custody of the source. Warden still produces a round's record; Anamnesis does not produce one. Dokimasia compiles a frontend's routes, actions and guards into a coverage denominator and reconciles a reviewed workbook against it, so every scoped item carries one disposition; it executes nothing, the harness that runs a release belongs to the application's own repository, and where Anamnesis keeps custody of findings and their remedies, Dokimasia records only which behaviour has no reviewed oracle. Sapheneia shapes the agent's replies for AuDHD readers and has one bounded operation for durable audit, issue, and comment prose. The same operation governs every other record its agents write down, from a pull request to a repository document or a commit message, and cuts each to the shortest form its evidence allows. It does not change another skill's facts or gates. Brevitas controls the volume and structure of engineering prose after vocabulary and register passes. If a request crosses one of those boundaries, hand it to the named sibling rather than broadening the selected skill.

Read the full file on GitHub · 391 lines

Changes

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.

  1. today Changed · +15 lines · +179 tokens per session 4a4d2db22e4b
  2. 2d ago Changed · +35 lines · +477 tokens per session 7808a73fa5f2
  3. 4d ago Changed · +77 lines · +1,006 tokens per session 766362dc0892
  4. 8d ago First seen · 264 lines · 3,115 tokens per session scan A 4d0a2fe794bc

Subscribe to this mod's changes

skills AGENTS.md is an instructions file published in the GitHub repository wildcat-finance/skills (12 stars, last pushed today), licensed Apache-2.0. It adds 4,777 tokens to every session, about $0.0239 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 2026-08-30.

Related

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.

vercel/next.js · 7,296 tokens

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.

openai/codex · 5,153 tokens

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).

microsoft/vscode · 6,785 tokens

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).

microsoft/vscode · 5,001 tokens

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.

langchain-ai/langchain · 4,469 tokens

spec-kit AGENTS.md

AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,104 tokens