Agent Substrate is a runtime system for running many isolated applications or AI agents on shared computing infrastructure. It manages agent sandboxes throughout their lifecycle, places them on available workers, and routes traffic to them while using technologies such as Kubernetes, microVMs, and gVisor. Its catalogue skills and instructions support operating and developing workflows around this agent runtime.
Borrowing it
Nothing to install: this file belongs to agent-substrate/substrate. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/agent-substrate/substrate/main/.agents/skills/agents-md/SKILL.mdgit clone --depth 1 https://github.com/agent-substrate/substrateWrote 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/skills/agent-substrate/substrate/agents-md)<a href="https://agentmods.dev/skills/agent-substrate/substrate/agents-md"><img src="https://agentmods.dev/badge/skills/agent-substrate/substrate/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.
<a href="https://agentmods.dev/skills/agent-substrate/substrate/agents-md"><img src="https://agentmods.dev/badge/skills/agent-substrate/substrate/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 5.1 | $0.00012 | $0.00545 |
| Opus 5.5 | $0.00005 | $0.00218 |
| Sonnet 5.5 | $0.00002 | $0.00109 |
| Haiku 4.5 | $0.00001 | $0.00055 |
Grade A, and why
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 2mo 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 — 72 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Generate or update AGENTS.md files in key folders of the project. Ensure the target AGENTS.md file accurately reflects the current status of the code.
Key Requirements
There must always be an AGENTS.md file in the project root. The AGENTS.md file at the root of the project may include the following sections:
- Project overview
- Build and test commands
- Code style guidelines
- Testing instructions
- Security considerations
AGENTS.md files in subfolders may be even more concise and specific to those subfolders.
Maximize AGENTS.md Performance
To maximize agent performance and token efficiency, adhere to the following:
Progressive Disclosure
Each AGENTS.md should be short. Approx. 100-150 lines total. Link to a handful of relevant, focused reference documents in the same repo, as needed.
Modularity
While only one AGENTS.md is allowed per folder, you can create an AGENTS.md file in any folder. Rather than putting everything in the root AGENTS.md, move highly specific advice to AGENTS.md in the most relevant folders of the project.
References
When documentation that would have helped you is missing, add references in the docs/ folder found at the project root. When these get out of date, update them.
Workflows
When references describe workflows, the tasks should be described as numbered steps. For example:
- Get the beep from the boop.
- Then bop it.
Decisions
When there are multiple ways of doing something in the codebase, use a decision table like the following:
| Question | Boop | Bop |
|---|---|---|
| Is foo a bar baz? | ✅ | |
| Is foo a quux? | ✅ |
Real Code Examples
If you think they are particularly good examples, include snippets from actual code in the repo, up to approx. 10 lines in length. These should only be the most relevant examples and representative of patterns that would be good to repeat in future code. Their purpose is to improve code reuse by agents, rather than each agent reinventing the wheel.
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.
- 2mo ago First seen · 72 lines · 12 tokens per session scan A b9effdec783f
agents-md is a skill published in the GitHub repository agent-substrate/substrate (5,957 stars, last pushed today), licensed Apache-2.0. It adds 12 tokens to every session and 545 once invoked, about $0.0000 per session on Opus 5.5. 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 skills, from other repositories
agent-engineering-expert
Build LLM agents that use tools safely: tool design, the agent loop, memory, MCP servers, multi-agent orchestration, sandboxing and prompt-injection defence. Use when the user mentions AI agents, tool use or function calling, MCP or Model Context Protocol, autonomous workflows, multi-agent systems, LangChain or…
background-remove
Remove backgrounds from portrait and general images, outputting PNGs with transparent backgrounds.
transcribe
Turn call recordings (mp3/audio) into speaker-labeled, timestamped transcripts (Markdown + JSON), fully local — faster-whisper large-v3 ASR + pyannote speaker diarization on the GPU.
scratch
Catch-all fallback for a task that matches no specialized workflow. Gives general input/output a home under the same write-boundary, logging, and human gates as every other workflow.
file-organizer
Analyze a messy folder and propose a reorganization. Produces a reviewable PLAN first; only applies changes after explicit user approval. Never deletes.
ocr-folder
Make a folder of scanned or image-only PDFs text-searchable. Per-page routing keeps existing text layers and OCRs only image pages; a content-hash cache skips unchanged files on re-runs. Pure pip — no system OCR binary needed.