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/rebeccapowell/agent-atlas/copilot-instructionsgit clone --depth 1 https://github.com/rebeccapowell/agent-atlasWrote 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/rebeccapowell/agent-atlas/copilot-instructions)<a href="https://agentmods.dev/instructions/rebeccapowell/agent-atlas/copilot-instructions"><img src="https://agentmods.dev/badge/instructions/rebeccapowell/agent-atlas/copilot-instructions.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 | $0.05080 | $0.05080 |
| Opus 5 | $0.02540 | $0.02540 |
| Sonnet 5 | $0.01016 | $0.01016 |
| Haiku 4.5 | $0.00508 | $0.00508 |
Grade A, and why
agent-atlas copilot-instructions.md scanned grade A with 1 finding 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 5d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -sf http://localhost:5063/healthz # should return "Healthy" How it starts
The opening of the file, as written. The whole thing — 410 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Copilot Instructions for Agent Atlas
Project Overview
Agent Atlas is an enterprise MCP (Model Context Protocol) gateway built with ASP.NET Core 10 and .NET Aspire. It turns a messy internal API estate into a governed, searchable catalog of MCP tools that both humans and AI agents can discover and invoke safely.
- Catalog: Reads OpenAPI specs from a GitOps data-plane repository. Teams annotate operations with
x-mcpvendor extensions to mark them as agent tools. - MCP Server: Exposes three MCP tools —
SearchTools,DescribeTool,ExecutePlan— via Streamable HTTP transport at/mcp. - Execution Engine: Interprets a JSON plan DSL (
call,foreach,if,returnsteps) and proxies requests to downstream APIs, forwarding the caller's JWT. - React UI: A read-only capability map for developers and operators (served from
wwwroot).
Repository Layout
src/
Atlas.AppHost/ # .NET Aspire orchestration host (local dev entry point)
Atlas.Host/ # Main service: MCP server, catalog API, execution engine, React UI
SampleApi.ToolEnabled/ # Demo API whose operations ARE registered as MCP tools
SampleApi.NotToolEnabled/ # Demo API intentionally NOT registered as tools
catalog/ # GitOps data-plane: catalog.yaml, apis/, policies/
docs/ # Architecture and deployment documentation
helm/ # Kubernetes / Helm chart
Tech Stack
- .NET 10 / ASP.NET Core 10
- .NET Aspire 13.x (orchestration, service discovery)
- ModelContextProtocol.AspNetCore 1.0.0 (MCP SDK)
- Keycloak (OIDC provider for local dev, via Docker)
- React (UI, pre-built into
wwwroot) - YamlDotNet (catalog YAML parsing)
- Microsoft.OpenApi.Readers (OpenAPI spec loading)
Build and Run
IMPORTANT — trust the HTTPS dev certificate first. Aspire uses a local HTTPS certificate for service-to-service communication via DCP (its internal process manager). If the certificate is not trusted, the DCP handshake times out after 20 seconds and the AppHost fails to start. Always run the three cert commands below once before the first
aspire runin any new environment.
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.
- 5d ago First seen · 410 lines · 5,080 tokens per session scan A 2b6da4451f0d
agent-atlas copilot-instructions.md is an instructions file published in the GitHub repository rebeccapowell/agent-atlas (2 stars, last pushed 6mo ago), licensed MIT. It adds 5,080 tokens to every session, about $0.0254 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
microsandbox AGENTS.md
AGENTS.md instructions for superradcompany/microsandbox, covering agents.md, scope, project map, design principles and backward compatibility review.
MultiGen CLAUDE.md
Instructions for lingyuanli/MultiGen, covering claude.md, architecture, development commands, full stack (docker) and api (local).
livecontext-ce AGENTS.md
AGENTS.md instructions for livecontext-ai/livecontext-ce, covering project guidelines (ai coding agents), before writing any code, architecture, database migrations and inter-service communication.
terfyn AGENTS.md
Instructions for LAA-Software-Engineering/terfyn, covering terfyn — agent instructions, layout, commands, conventions and skills.
n8n-starter CLAUDE.md
Instructions for gabry-ts/n8n-starter, covering claude.md, project overview, repository structure, workflow files and credential management.
terfyn CLAUDE.md
Instructions for LAA-Software-Engineering/terfyn, a project described as: A statically analyzable execution platform for nondeterministic programs, with reviewable authority and reproducible deployment state.