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/pacnpal/mcpelevator/agents-mdgit clone --depth 1 https://github.com/pacnpal/mcpelevatorWhat 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 | $0.01557 | $0.01557 |
| Opus 5 | $0.00779 | $0.00779 |
| Sonnet 5 | $0.00311 | $0.00311 |
| Haiku 4.5 | $0.00156 | $0.00156 |
Grade A, and why
mcpelevator 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 3d 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 — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
mcpelevator — self-hosted control plane that elevates stdio MCP servers into authenticated HTTP endpoints
One container runs stdio (or already-remote) MCP servers and exposes each as a remote
Streamable-HTTP endpoint (/s/<slug>/mcp) with a SvelteKit UI, process supervision, and auth.
Commands
Run backend commands from backend/, frontend commands from frontend/. The Makefile
wraps the common ones.
- Backend dev (autoreload, http://127.0.0.1:8080):
cd backend && uv run uvicorn app.main:app --reload --port 8080—make dev-backend - Frontend dev (HMR, http://localhost:5173, proxies /api and /s to :8080):
cd frontend && npm install && npm run dev—make dev-frontend - Build SPA into
frontend/build:cd frontend && npm ci && npm run build—make build - Backend tests:
cd backend && uv run pytest -q—make test - Frontend tests:
cd frontend && npm run test(vitest) - Frontend typecheck:
cd frontend && npm run check(svelte-check) - Refresh the uv lockfile:
cd backend && uv lock—make lock - Build + run everything in Docker:
docker compose up --build—make docker
Architecture
One FastAPI process serves three surfaces in a single port (backend/app/main.py):
/api/* control plane, /s/<slug>/mcp reverse proxy, and the built SPA as a catch-all mount.
- Desired-state reconciliation. SQLite is the source of truth. A background supervisor task
(
supervisor/) converges running processes to the desired state (Kubernetes-style), so the system is idempotent and survives restarts. - One bridge process per enabled server (
bridge/,runners/): each runs its own uvicorn on a loopback port hosting a FastMCP proxy of the stdio command (or an upstream HTTP/SSE URL), fault-isolated with a real PID and logs. - Upstream OAuth (
auth/oauth_store.py,auth/oauth_flow.py): aremoteserver can authenticate to its upstream via OAuth instead of staticenvheaders. The interactive authorization-code grant (DCR + PKCE) runs in the control plane (/api/servers/{id}/oauth/authorize→ public/api/oauth/callback, anchored on the OAuthstate) using the MCP SDK'sOAuthClientProvider; tokens land in a0600file store (<data_dir>/oauth/<id>.json) shared with the bridge, which reads them and auto-refreshes. Tokens live off the DB, so authenticating never re-hashes the row or bounces the bridge. - Runners (
runners/):npx,uvx,command,remote(proxy an already-remote MCP URL), anddocker(image-packaged servers — opt-in + root-equivalent behind thedocker_runnersetting). Each runner is a pureServer -> ProcessSpecbuilder;dockerstores the canonical image+container-args+env shape and synthesizes a hardeneddocker run(the bridge scrubs the child env for docker units so a-e KEYpassthrough can't reach the control plane's secrets). - Groups (
groups/): a SQLite-backed registry maps each group name to"*"or an ordered list of server IDs and serves it at/g/<name>/mcp;allis an ordinary, unreserved group name.GroupHubconverges one namespaced FastMCP bundle per group from the running supervisor topology after each reconcile. Auth reusesenforce()with syntheticgroup:<name>identities, so bearer tokens may be scoped to that group orall; members whose effective auth is stricter than the group are excluded to prevent an auth downgrade. - Auth (
auth/): two independent layers per request — a Host/Origin allowlist middleware (DNS-rebinding defense) plus pluggable per-server bearer auth on/sand control-plane bearer auth that gates the sensitive/apirouters only when enforcement is on (defaultauto: when the box is exposed off-host);/api/health*and/api/auth/statusstay public. See README "Security" for the full model. - Catalog (
catalog/): backend proxies public MCP directories (official registry, Glama) into reviewable launch specs; the SPA stays same-origin. - Frontend (
frontend/src/): SvelteKit (Svelte 5) SPA,adapter-static, no SSR — rendered entirely in the browser, served by the backend catch-all.
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.
- 3d ago First seen · 83 lines · 1,557 tokens per session scan A 4426834821e2
mcpelevator AGENTS.md is an instructions file published in the GitHub repository pacnpal/mcpelevator (5 stars, last pushed 11d ago), licensed MIT. It adds 1,557 tokens to every session, about $0.0078 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
moira AGENTS.md
Instructions for moira-mcp/moira, covering agents.md, what moira is, repository layout, build & run (fresh clone) and image: line and uncomment the build: block, then.
moira CLAUDE.md
Instructions for moira-mcp/moira, a project described as: Agent Workflow Engine for AI agents over MCP — per-step directives, completion conditions, and JSON-Schema validation. Self-hostable (Apache-2.0).
ODS CLAUDE.md
Claude Code instructions for Osmantic/ODS, covering claude.md, project overview, repository structure, build & development commands and linting and validation.
meme-search CLAUDE.md
Instructions for neonwatty/meme-search, covering claude.md, 📁 file organization guidelines, 🤖 task agent usage guidelines, when to use task agents and documentation lookup pattern.
ProductFlow AGENTS.md
Instructions for yuqie6/ProductFlow, covering trellis instructions, repository guidelines, project structure & module organization, build, test, and development commands and coding style & naming conventions.
homelab-monitor CLAUDE.md
Instructions for SikamikanikoBG/homelab-monitor, covering claude.md and conventions.