toughtongue-skills: Instructions file for Claude Code

AGENTS.md

toughtongue-skills AGENTS.md is an instructions file for Claude Code, Codex, OpenCode from tough-tongue/toughtongue-skills. It costs 2,464 tokens per session, scanned A, original, MIT.

A set of contribution rules for toughtongue-skills, a public repository containing agent skills and plugin settings for the Tough Tongue AI MCP server. MCP is a standard way for agents to call external tools.

In plain words
What is it for?
Use it when adding or updating skills, MCP tool references, authentication settings, documentation, or releases in this repository.
Why use it?
It helps keep published instructions compatible with the server, free of secrets, and short enough for an agent to use effectively.

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: reads .claude/ paths; mentions Claude Code; mentions Codex.

This is tough-tongue/toughtongue-skills's own configuration. It tells Claude Code, Codex and OpenCode how to work on toughtongue-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 toughtongue-skills configures →

Reuse

Borrowing it

Nothing to install: this file belongs to tough-tongue/toughtongue-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/tough-tongue/toughtongue-skills/main/AGENTS.md
Clone the repo
git clone --depth 1 https://github.com/tough-tongue/toughtongue-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 toughtongue-skills AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/tough-tongue/toughtongue-skills/agents-md/github.svg)](https://agentmods.dev/instructions/tough-tongue/toughtongue-skills/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/tough-tongue/toughtongue-skills/agents-md"><img src="https://agentmods.dev/badge/instructions/tough-tongue/toughtongue-skills/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.

agentmods 80×15 button for toughtongue-skills AGENTS.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/tough-tongue/toughtongue-skills/agents-md"><img src="https://agentmods.dev/badge/instructions/tough-tongue/toughtongue-skills/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 2,464 This file is loaded in full into every session.
When invoked 2,464 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 $0.02464 $0.02464
Opus 5 $0.01232 $0.01232
Sonnet 5 $0.00493 $0.00493
Haiku 4.5 $0.00246 $0.00246

Measured 3d ago against content hash 845fa7eb6d5c, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

toughtongue-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 3d 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.

AGENTS.md · 144 lines

How it starts

The opening of the file, as written. The whole thing — 144 lines — stays where its author put it; the contents beside it link to each section on GitHub.

toughtongue-skills — Agent Guide

Skills + plugin manifests for the Tough Tongue AI MCP server. This repo is public and is installed directly into end users' agents — every word ships.

Rules

  • MCP tool names are a contract. Skills may only reference tools that the hosted MCP server exposes (full catalog in MCP.md). If the server catalog changes, MCP.md and the skills change in the same PR.
  • Skills are MCP-first. Every workflow ends in an MCP tool call (create_scenario, update_scenario, list_sessions, ...) — never in "write a file to disk" or references to internal Tough Tongue AI repos, paths, or CLIs.
  • No secrets. Authentication is OAuth-first: the shipped MCP configs (.mcp.json / mcp.json) carry no credentials and clients run the browser OAuth flow. The only credential ever named is the TTAI_PAT environment variable (headless/CI fallback), referenced by name only. Never hardcode tokens, user IDs, or org IDs.
  • Token discipline in SKILL.md files. Skills are loaded into agent context; keep SKILL.md under ~250 lines and push depth into references/ files that agents load on demand.
  • Public bar. No internal jargon, no unfinished sections, no hallucinated features. Terminology: "Tough Tongue AI" (two words, matching the public brand at toughtongueai.com), not "ToughTongue AI". Positioning: the platform for tough conversations, with a dual storyline — some the AI takes (voice agents that call, demo, screen, book), others you nail (realistic roleplay for negotiations, interviews, coaching). Do not describe the platform as only roleplay/training, and do not drop the practice half either — lead with the automate + practice pairing.

Structure

Skills are layered. Load a lower layer when a higher one needs it — do not copy field tables or OAuth dumps into workflow skills.

  • Layer 0 — skills/ttai-agent/intermediate intelligence layer. It turns human intent into a scoped, capability-aware decision across the Tough Tongue AI datastore: control entities, resources, scenario shape, runtime channels, situation recipes (kb/), and MCP guidance (features/mcp/). It is consumer-neutral: coding agents and web conversational plugins share the same knowledge; their adapters execute. Each subdirectory has an index.md.
  • Layer 1 — scenario-maker, session-analyst, browser-demo-builderwhen. Workflows that load ttai-agent.

Read the full file on GitHub · 144 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. 3d ago Changed · +21 lines · +380 tokens per session 845fa7eb6d5c
  2. 12d ago First seen · 123 lines · 2,084 tokens per session scan A 98bd9dfb8b4e

Subscribe to this mod's changes

toughtongue-skills AGENTS.md is an instructions file published in the GitHub repository tough-tongue/toughtongue-skills (7 stars, last pushed today), licensed MIT. It adds 2,464 tokens to every session, about $0.0123 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.

Related

Other instructions, from other repositories