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/acnlabs/metaspec/agents-mdgit clone --depth 1 https://github.com/acnlabs/MetaSpecWhat 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 | $0.12090 | $0.12090 |
| Opus 5 | $0.06045 | $0.06045 |
| Sonnet 5 | $0.02418 | $0.02418 |
| Haiku 4.5 | $0.01209 | $0.01209 |
Grade A, and why
MetaSpec 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 2d 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 — 1,422 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Meta-Spec AI Agent Guide
For AI Assistants: This document provides guidance on using Meta-Spec to generate Spec-Driven X (SD-X) speckits.
🎯 Your Role
You are helping a developer create speckits (spec-driven toolkits for AI agents) using MetaSpec.
MetaSpec is a meta-specification framework that generates complete Spec-Driven X (SD-X) speckits:
- SD-Development - Spec-driven development
- SD-Design - Spec-driven design systems
- SD-Testing - Spec-driven testing frameworks
- SD-Documentation - Spec-driven documentation
- SD-Operations - Spec-driven operations
- SD-X - Spec-driven generation for any domain
Key principle: MetaSpec generates production-ready speckits with CLI, parser, validator, templates, and AI agent support
💡 Core Concept: Meta-Specification Framework
MetaSpec is a meta-specification framework - it uses specifications to generate specification toolkits.
Meta-Specification (MetaSpec)
↓
Defines how to define specifications
↓
Generates Speckits (Spec-driven Toolkits)
↓
Speckits carry domain specifications
↓
Used to develop domain projects
Key Insights
- Meta-level abstraction - MetaSpec defines how to create specification systems
- Generative framework - Generates complete speckits from meta-specifications
- Recursive architecture - MetaSpec uses SDD to develop itself, generates speckits that use SDD
- Specification-centric - Domain specs are the source of truth, speckits are carriers
Example Flow:
MetaSpec (meta-spec)
↓ generates
MCP-Spec-Kit (carries MCP specification spec)
↓ used by
Developer (validates MCP servers against spec)
This means: MetaSpec is not just "spec-driven development" - it's a framework that generates speckits (spec-driven toolkits) for any domain.
See Recommended Practice: SDS + SDD Separation in MetaSpec Commands section for how speckits separate domain specifications from toolkit implementation.
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.
- 2d ago First seen · 1,422 lines · 12,090 tokens per session scan A 81944f8fce3c
MetaSpec AGENTS.md is an instructions file published in the GitHub repository acnlabs/MetaSpec (51 stars, last pushed 9mo ago), licensed MIT. It adds 12,090 tokens to every session, about $0.0605 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.
Other instructions, from other repositories
novel-writer copilot-instructions.md
Instructions for wordflowlab/novel-writer, covering quick context for ai coding agents, high-level architecture notes, common developer workflows (how to run / build / test), project-specific conventions and gotchas (do not invent) and what to change when adding or updating an agent.
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.
spec-driven-steroids AGENTS.md
Instructions for lindoelio/spec-driven-steroids, covering agents.md, project identity, technology stack, essential commands and agent constraints.
leanspec AGENTS.md
Instructions for codervisor/leanspec, covering ai agent instructions, project: leanspec, skills, spec-driven development (github-issue specs) and development & tooling.
AmbyKit AGENTS.md
Instructions for ambystechcom/AmbyKit, covering agents.md — ambykit, what ambykit is, golden rules, project structure and build & test.
ai-workflow-kit AGENTS.md
Instructions for silver2dream/ai-workflow-kit, covering agents.md (worker agent guide), role: senior engineer (worker), must-read (before any work), non-negotiable hard rules and 0. use existing architecture & do not reinvent.