Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/mulkatz/anvil/claude-mdgit clone --depth 1 https://github.com/mulkatz/anvilWrote 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.
[](https://agentmods.dev/instructions/mulkatz/anvil/claude-md)<a href="https://agentmods.dev/instructions/mulkatz/anvil/claude-md"><img src="https://agentmods.dev/badge/instructions/mulkatz/anvil/claude-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.01139 | $0.01139 |
| Opus 5 | $0.00570 | $0.00570 |
| Sonnet 5 | $0.00228 | $0.00228 |
| Haiku 4.5 | $0.00114 | $0.00114 |
Grade A, and why
anvil CLAUDE.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.
How it starts
The opening of the file, as written. The whole thing — 93 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Anvil — Adversarial Thinking Plugin
What this is
A Claude Code plugin that enables adversarial thinking through structured debates. Uses stop hook orchestration to rotate through Advocate → Critic → Synthesizer phases, each with distinct role prompts.
Architecture
- No TypeScript runtime — the plugin is shell scripts + markdown prompts
bunis used as package manager only (scripts, formatting)- The prompts in
prompts/ARE the product - Stop hook (
hooks/stop-hook.sh) is the orchestrator — it manages the state machine - State lives in
.claude/anvil-state.local.md(YAML frontmatter + markdown transcript)
Key files
hooks/stop-hook.sh— Core state machine and prompt routingscripts/setup-anvil.sh— Argument parsing, validation, state file creationprompts/{advocate,critic,synthesizer}.md— Role-specific instructionsprompts/modes/{analyst,philosopher,devils-advocate}.md— Mode-specific tonecommands/anvil.md— Entry point command
Documentation
When a change affects user-facing behavior (new flags, new features, changed defaults, new
modes/frameworks/personas), update README.md accordingly:
- New CLI flag: Add to the options table and, if significant, add a section under Features
- New mode/framework/persona preset: Add to the relevant Features section with an example
- Changed behavior: Update any affected descriptions, examples, or the debate flow diagram
- New script/file: Update the Architecture tree if it's a top-level addition
Conventions
- ADRs in
docs/adr/for architectural decisions - State file uses
.local.mdsuffix (gitignored by Claude Code) - All shell scripts use
set -euo pipefail - Frontmatter parsing with
sed, transcript manipulation withawk - Atomic file updates via temp file +
mv
Testing
Every change to setup-anvil.sh or stop-hook.sh MUST include corresponding test updates. Run
bun run check (shellcheck + all bats tests) before committing — it must pass.
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.
- 3d ago First seen · 93 lines · 1,139 tokens per session scan A a5dac10aefd9
anvil CLAUDE.md is an instructions file published in the GitHub repository mulkatz/anvil (1 stars, last pushed 6mo ago), licensed MIT. It adds 1,139 tokens to every session, about $0.0057 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.
Other instructions, from other repositories
cc-thinking-skills CLAUDE.md
Instructions for tjboudreaux/cc-thinking-skills: Follow AGENTS.md for repository workflow.
cc-thinking-skills AGENTS.md
Instructions for tjboudreaux/cc-thinking-skills, covering agent guidelines, first principles, skill authoring & eval standards, default workflow and verification.
platform-skills copilot-instructions.md
Copilot instructions for nitinjain999/platform-skills, covering platform engineering instructions for github copilot, version: 1.40.0, source: https://github.com/nitinjain999/platform-skills, response contract and core principles.
archsight-cognition CLAUDE.md
Instructions for ArchSightLabs/archsight-cognition, covering claude.md, 必读入口 and claude 特别说明.
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).
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.