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/kolsetu-opensource/elba-mcp-server/claude-mdgit clone --depth 1 https://github.com/Kolsetu-Opensource/elba-mcp-serverWrote 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/kolsetu-opensource/elba-mcp-server/claude-md)<a href="https://agentmods.dev/instructions/kolsetu-opensource/elba-mcp-server/claude-md"><img src="https://agentmods.dev/badge/instructions/kolsetu-opensource/elba-mcp-server/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 5.1 | $0.00479 | $0.00479 |
| Opus 5 | $0.00239 | $0.00239 |
| Sonnet 5 | $0.00096 | $0.00096 |
| Haiku 4.5 | $0.00048 | $0.00048 |
Grade A, and why
elba-mcp-server 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 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.
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.
What it actually says
CLAUDE.md
This file provides context for AI assistants working on this codebase.
Project Overview
This is an open-source MCP (Model Context Protocol) server for Elba, a voice AI agent platform. It exposes Elba's agent lifecycle as MCP tools callable from any MCP-compatible client (Claude Code, Cursor, VS Code, etc.).
Current Status
Phase 1 tools are implemented and wired into the server:
list_agents— lists all agents in the organizationlist_voices— lists available voicesupsert_agent— creates or updates a voice AI agent
Auth validation (src/auth.ts) checks token format (must start with elba_). Full token validation happens server-side via the Elba API.
Architecture
- Thin client pattern: The MCP server is a thin wrapper around Elba's backend API. All business logic lives in the Elba platform — this server just translates MCP tool calls into API requests.
- Authentication: Uses integration tokens (
ELBA_INTEGRATION_TOKENenv var) validated against the Elba backend on every tool call. - Transport: stdio (local) by default. The server is started by the MCP client process.
Key Files
src/index.ts— CLI entry point (shebang, starts server)src/server.ts— MCP server initialization, tool registrationsrc/auth.ts— Token validation against Elba backendsrc/tools/— One file per MCP tool__tests__/— Mirrors src/ structure, uses vitest
Commands
npm run build— Compile TypeScript to dist/npm run dev— Run server in development mode (tsx)npm test— Run tests (vitest)npm run lint— ESLintnpm run typecheck— TypeScript type checkingnpx changeset— Create a changeset for version bumping
Conventions
- ESM only (
"type": "module"in package.json) - Imports use
.jsextensions (NodeNext module resolution) - One tool per file in
src/tools/ - Tests mirror source structure in
__tests__/ - All tools validate auth before executing
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 · 48 lines · 479 tokens per session scan A d7153e93882c
elba-mcp-server CLAUDE.md is an instructions file published in the GitHub repository Kolsetu-Opensource/elba-mcp-server (3 stars, last pushed 2d ago), licensed MIT. It adds 479 tokens to every session, about $0.0024 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
mcp-ts-core AGENTS.md
AGENTS.md instructions for cyanheads/mcp-ts-core, covering developer protocol, consumers, core rules, exports reference and import conventions.
wanderlog-mcp CLAUDE.md
Claude Code instructions for shaikhspeare/wanderlog-mcp, covering wanderdog — agent instructions, read before making changes, keep the kanban in sync, invariants (from docs/architecture.md — do not violate) and dev commands.
webmcp-react AGENTS.md
AGENTS.md instructions for agentcathq/webmcp-react, covering agents.md, reference, commands, project structure and architecture decisions — don't "fix" these.
better-godot-mcp CLAUDE.md
Claude Code instructions for n24q02m/better-godot-mcp, covering claude.md - better-godot-mcp, commands, setup, lint & type check and fix.
better-godot-mcp AGENTS.md
AGENTS.md instructions for n24q02m/better-godot-mcp, covering agents.md - better-godot-mcp, build / lint / test commands, run a single test file, run a single test by name and mise shortcuts.
now-sdk-ext-mcp CLAUDE.md
Instructions for sonisoft-cnanda/now-sdk-ext-mcp, covering now-sdk-ext-mcp, project overview, architecture, directory structure and sibling projects.