Observed Signal · May 2, 2026 · Technical Release · Source: DEV Community · Impact: 4/5 · Sentiment: Neutral
AI Instruction Split: AGENTS.md, SKILL.md, DESIGN.md
The article describes a growing three-layer standard for instructing AI agents: AGENTS.md for overall agent behavior and boundaries, SKILL.md for reusable task procedures (used by Anthropic's Claude Skills and the Agent Skills standard), and DESIGN.md — a Google Labs design-spec format released in April 2026 that combines machine-readable design tokens (YAML) with human-readable intent and ships with a CLI validator (npx @google/design.md lint). The author argues these formats separate verifiable rules (tokens, audits, structural checks) from judgment-based guidance (tone, stance), and situates the split alongside Spec-Driven Development (SDD) workflows (Kiro, GitHub Spec Kit). The three-layer approach is presented as complementary to SDD and intended for incremental adoption where verification adds value.
A technical specification and validator released by Google Labs (a major platform) can standardize machine-readable design tokens and agent instruction formats, influencing developer tooling, UI generation by agents (Stitch), and cross-vendor agent interoperability—impacting how teams operationalize AI agents and specs.
Track Anthropic Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Google Labs published the DESIGN.md specification on April 10, 2026 and provides a reference implementation used by Stitch.
- DESIGN.md files contain machine-readable design tokens in YAML and human-readable design intent in Markdown, and Google ships a CLI validator: npx @google/design.md lint.
- AGENTS.md emerged as an industry standard since 2025 (jointly developed by OpenAI, Google, Sourcegraph, Cursor, and Factory) and was donated to the Linux Foundation in December 2025.
- SKILL.md is the file format standardized by agentskills.io and underpins Anthropic's Claude Skills; the same SKILL.md can run across Claude Code, Claude.ai and other Agent SDKs.
- The DESIGN.md repo had over 11,000 stars on GitHub as of early May 2026 and the spec includes linting checks such as token integrity and WCAG contrast ratios.
Connected Companies & Entities
4 Entities mappedOntology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
AgentSkills: Teach AI Agents How to Execute Tasks
The article describes a gap in many LLM-based agent applications: agents often know what to do but not how to do it reliably. It introduces AgentSkills (aka Procedure Skills) — self-contained, structured playbooks (commonly formatted as SKILL.md) that bundle YAML frontmatter, step-by-step execution instructions, small automation scripts, domain resources, and output templates. The author explains why embedding full procedures in large system prompts fails (fragility, token waste, inconsistency) and advocates progressive disclosure: a discovery phase that loads only skill names/descriptions and an activation phase that loads full skill assets when a match occurs. The piece gives design principles for effective skills (imperative language, explicit failure states, small composable units) and explains when skills materially improve agent reliability and cost-efficiency. Published May 6, 2026 by Sreeni Ramadorai on DEV Community.
DESIGN.md vs tokens.json vs Figma for AI Agents
A developer post (PromptMaster) published on 2026-06-27 compares three approaches for giving AI agents design context: tokens.json, prose README files, and Figma links — and argues DESIGN.md combines the strengths of all three. The article states tokens.json provides exact values but cannot express application rules; prose README files can express rules but lack structured, machine-readable tokens; and Figma is designed for humans and is unreadable directly by coding agents. DESIGN.md is presented as a single, versioned file that provides structured values, expressible rules, machine readability, and persistence, and can export to Tailwind and the W3C DTCG standard via a CLI (npx @google/design.md). The post includes links to a free cheat sheet and a paid full guide on Gumroad.
AI Coding Agents Need Product Design Context
The article argues that AI coding agents can fully read a codebase but still produce generic, off-brand outputs because much of a product’s identity—tone, visual language, interaction principles and positioning—does not live in code. The author, Gregory Muryn‑Mukha, describes seven distinct types of knowledge that form a product’s “design context” and documents a practical solution: a Claude Code skill directory (.claude/skills/<product>-context/) that organizes progressive, task‑specific files (SKILL.md router, design.md, quickref, references) to load the right design and engineering constraints into an agent. The piece details how the skill was iteratively built, common failure modes, use cases (Figma→code lookup, component authoring, product brainstorming, landing‑page copy), and maintenance trade-offs such as drift and the need for periodic review.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.