# MESA — full corpus
Every document in the idss-mesa.github.io OKF v0.2 bundle. Each begins with its canonical URL followed by its original Markdown, OKF frontmatter included. Relative links resolve against that document's URL.
---8<--- https://idss-mesa.github.io/okf/index.md
---
okf_version: "0.2"
title: "MESA knowledge bundle"
description: "Open Knowledge Format (OKF) bundle indexing the MESA (Multidisciplinary Environment for Scientific Advancement) website, documentation, and software for human and agentic AI consumption"
---
# idss-mesa.github.io
This is the root of an [Open Knowledge Format (OKF) v0.2](https://github.com/GoogleCloudPlatform/knowledge-catalog/blob/main/okf/SPEC.md)
bundle describing MESA — the Multidisciplinary Environment for Scientific
Advancement, an open-source, agentic AI platform for scientific data (NSF
IDSS Category II, University of New Mexico). Each page of
has a concept here that is its Markdown twin.
All content on this origin, including the
[/docs/](https://idss-mesa.github.io/docs/) and
[/neon-mcp/](https://idss-mesa.github.io/neon-mcp/) sub-sites, is open to AI
crawlers under . The agent guide is
[For AI agents](concepts/ai-agents.md); the whole bundle in one file is
. Related bundles:
and
.
# Site
* [MESA — Multidisciplinary Environment for Scientific Advancement](concepts/home.md) - MESA is an open-source, agentic AI platform that connects scientific data through curated, indexed metadata — a data lakehouse, a federated data mesh, and AI agents that speak both. NSF IDSS Category II project at the University of New Mexico
* [MESA Hiring](concepts/hiring.md) - MESA and CARC are hiring at the University of New Mexico — full-time positions at the Center for Advanced Research Computing in Albuquerque
* [For AI agents](concepts/ai-agents.md) - How agents and harnesses should read idss-mesa.github.io — robots.txt, llms.txt, Markdown twins with OKF v0.2 frontmatter, and trust signals — and why to call the MESA MCP servers for data
# Documentation
* [MESA MCP Stack Documentation](concepts/docs.md) - One-line install of the CyVerse MESA MCP stack (mesa-mcp, mesa-ducklake, irods-mcp-server, formation-mcp) for Claude Code, Codex CLI, Antigravity, and OpenCode
# Software
* [mesa-mcp](concepts/mesa-mcp.md) - MCP server bridging the CyVerse Data Store (iRODS) with OBO/OLS ontology AVU tools, DataCite metadata, and DuckLake metadata history
* [mesa-ducklake](concepts/mesa-ducklake.md) - AVU metadata-history library for MESA projects (DuckLake: Postgres/DuckDB catalog plus Parquet in iRODS)
* [irods-mcp-server](concepts/irods-mcp-server.md) - Reference MCP server for the CyVerse Data Store (iRODS), written in Go
* [formation-mcp](concepts/formation-mcp.md) - MCP server for the CyVerse Discovery Environment (Formation API): launch apps and manage analyses, written in Go
* [neon-mcp](concepts/neon-mcp.md) - MCP (2026-07-28) server for the NEON Data API: discover, check availability, list, download and cite NEON ecological data
* [mesa-sandbox](concepts/mesa-sandbox.md) - AI sandboxes for CyVerse VICE: a sudo-free multi-arch image family, sandbox profiles P0–P4, an OpenBao credential broker and egress proxy, and agent-sandbox pods for autonomous agents
* [MESA CLI](concepts/cli.md) - CyVerse Cloud Shell CLI with MESA tools
* [mesa-portal](concepts/mesa-portal.md) - Portal based on the CyVerse Terrain API
* [mesa-forge](concepts/mesa-forge.md) - A metadata extraction tool generation and management library
---8<--- https://idss-mesa.github.io/okf/concepts/home.md
---
type: Website
title: "MESA — Multidisciplinary Environment for Scientific Advancement"
description: "MESA is an open-source, agentic AI platform that connects scientific data through curated, indexed metadata — a data lakehouse, a federated data mesh, and AI agents that speak both. NSF IDSS Category II project at the University of New Mexico"
resource: https://idss-mesa.github.io/
tags: [site, mesa, agentic-ai, data-lakehouse, data-mesh, mcp, cyverse]
status: stable
generated: { by: "claude/opus-5", at: "2026-09-10T00:00:00Z" }
stale_after: "2027-03-10T00:00:00Z"
sources:
- id: landing-src
resource: https://github.com/idss-mesa/idss-mesa.github.io/blob/main/index.html
title: "MESA landing page source (index.html)"
author: "team:idss-mesa"
- id: nsf-award
resource: https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2632685
title: "NSF Award #2632685 — MESA"
author: "team:nsf"
- id: install-sh
resource: https://github.com/idss-mesa/docs/blob/main/install.sh
title: "MESA MCP stack installer (install.sh)"
author: "team:idss-mesa"
- id: openai-black-holes
resource: https://openai.com/index/creating-new-simulations-black-holes/
title: "Creating new simulations of black holes (OpenAI)"
author: "team:openai"
---
# MESA — Multidisciplinary Environment for Scientific Advancement
**Ask the question. MESA finds the data.**
MESA — the Multidisciplinary Environment for Scientific Advancement — is an
open-source, agentic AI platform that connects scientific data through
curated, indexed metadata. A data lakehouse, a federated data mesh, and AI
agents that speak both, running on national research infrastructure
today.[^landing-src]
This document is the Markdown twin of the landing page at
. Documentation lives at
; code at .
## Five integrated layers
| # | Layer | Technologies |
|---|-------|--------------|
| 01 | User interface | Laptop · Desktop · K8s container |
| 02 | Agentic AI | MCP · RAG · Agents · Reasoning · Sandboxing |
| 03 | Data management | Data Lakehouse (DuckLake) · Data Mesh (iRODS) |
| 04 | Compute | CyVerse · TACC · ACCESS-CI · NAIRR |
| 05 | Object storage | S3 · OSN · VAST · Lustre |
## Three innovation goals
1. **Open-source data lakehouse.** Analytical tables over open formats —
DuckLake, Apache Iceberg, Parquet — with time-travel versioning built in.
Query five petabytes like a database, and rewind any table to the state it
held when the paper was submitted.
2. **Federated iRODS data mesh.** One policy engine across the CyVerse Data
Store and its federated peers, with CILogon, Globus Auth, and ORCID
handling identity. The lakehouse and the mesh interoperate — your data
stays where it lives, and still shows up where you work.
3. **Agentic AI orchestration.** Self-hosted LLM serving on vLLM, RAG
pipelines over curated metadata, and an MCP server framework with
sandboxed, multi-agent orchestration. Agents that find, describe, and move
scientific data — with the guardrails written first.
## The stack, today
One line. Four servers. No credentials required.
MESA is not a promise — the MCP stack is public and installs now. One command
clones, builds, and registers four CyVerse servers in Claude Code (and in
Codex CLI, Antigravity, and OpenCode); anonymous public access to the Data
Store works out of the box.[^install-sh]
```bash
curl -fsSL https://raw.githubusercontent.com/idss-mesa/docs/main/install.sh | bash
```
| Server | Language | Role |
|--------|----------|------|
| [mesa-mcp](mesa-mcp.md) | Python | iRODS Data Store + OBO/OLS ontology metadata, DataCite, DuckLake history |
| [mesa-ducklake](mesa-ducklake.md) | Python | AVU metadata-history library backing mesa-mcp |
| [irods-mcp-server](irods-mcp-server.md) | Go | Reference iRODS Data Store server |
| [formation-mcp](formation-mcp.md) | Go | CyVerse Discovery Environment — launch apps, manage analyses |
### Featured connectors. Nothing to install.
The Data Store and Discovery Environment servers also run hosted, as remote
MCP connectors — no local build, no toolchain. In Claude Desktop or
claude.ai, open Settings → Connectors → Add custom connector, name the
connector, paste its endpoint URL, and authenticate when prompted. A
connector added to a claude.ai organization follows you everywhere Claude
runs: Claude Desktop, claude.ai, and Claude Code on the web.
| Connector | Server | What it opens |
|-----------|--------|---------------|
| cyverse-irods-shared | [irods-mcp-server](irods-mcp-server.md) | CyVerse Data Store — public collections under `/iplant/home/shared`, anonymous read |
| cyverse-formation | [formation-mcp](formation-mcp.md) | Discovery Environment — browse data, launch apps, track analyses |
Claude Code registers the same endpoints from the terminal (swap in the
current endpoint URLs from the [documentation](docs.md), which also
covers credentialed access, running either server locally in Claude Desktop
via `claude_desktop_config.json`, and self-hosting the endpoints):
```bash
claude mcp add --transport http irods https://IRODS-ENDPOINT/mcp
claude mcp add --transport http formation https://FORMATION-ENDPOINT/mcp
```
## The team
| Name | Role |
|------|------|
| Tyson L. Swetnam, Ph.D. | Principal Investigator · UNM |
| Arthur B. Maccabe, Ph.D. | Co-PI · Arizona site lead |
| David Ebert, Ph.D. | Co-PI · Chief AI & Data Science Officer · Arizona |
| Lei Cao, Ph.D. | Co-PI · Lakehouse lead · Arizona |
| Terrell Russell, Ph.D. | Co-PI · Data Mesh Lead · RENCI Site Lead |
| Illyoung Choi, Ph.D. | Research Software Engineer · Arizona |
| Kory Draughn | Chief Technologist · RENCI |
| Alan King | Senior Software Developer · RENCI |
| Justin James | Applications Engineer · RENCI |
| Sarah Roberts | Research Software Engineer · Arizona |
| Tony Edgin | Data Curator · Arizona |
| Edwin Skidmore | Cloud Orchestration / Infrastructure · CyVerse |
| Greg Chism, Ph.D. | EOT Lead · Senior Personnel · Arizona |
| Jingdi Chen, Ph.D. | 5G/6G Networking · Arizona |
| CK Chan, Ph.D. | Astronomer · Arizona |
| Wolfgang Jentner, Ph.D. | Human-Computer Interaction · Arizona |
| Junjia Guo | PhD Student · Arizona |
| Pengxin Wang | PhD Student · Arizona |
| Junyong Zhao | PhD Student · Arizona |
| Linan Zheng | PhD Student · Arizona |
MESA is hiring; see [MESA Hiring](hiring.md).
## News
- **MESA kicks off with NSF IDSS, July 2026.** The project team convenes with
the NSF Integrated Data Systems & Services program for the MESA kick-off
meeting — the University of New Mexico, the University of Arizona, and
UNC-CH/RENCI, setting the two-year prototype baseline in motion.
- **Creating new simulations of black holes, with OpenAI.** Co-investigator
CK Chan, Ph.D. worked with OpenAI on creating new simulations of black
holes — a look at agentic coding against Event Horizon Telescope
science.[^openai-black-holes]
## Funding
MESA is supported by the U.S. National Science Foundation under Award
#2632685 (Integrated Data Systems & Services, 2026–2029).[^nsf-award] Any
opinions, findings, and conclusions are those of the authors and do not
necessarily reflect the views of the NSF. Partner institutions: University of
New Mexico, University of Arizona, UNC-Chapel Hill / RENCI.
Agents: see [For AI agents](ai-agents.md) for how to read this
origin, and connect to the MCP servers above for the data itself.
[^landing-src]: MESA landing page source (index.html)
[^nsf-award]: NSF Award #2632685 — MESA
[^install-sh]: MESA MCP stack installer (install.sh)
[^openai-black-holes]: Creating new simulations of black holes (OpenAI)
---8<--- https://idss-mesa.github.io/okf/concepts/hiring.md
---
type: Webpage
title: "MESA Hiring"
description: "MESA and CARC are hiring at the University of New Mexico — full-time positions at the Center for Advanced Research Computing in Albuquerque"
resource: https://idss-mesa.github.io/hiring.html
tags: [site, hiring, jobs, unm, carc]
status: stable
generated: { by: "claude/opus-5", at: "2026-09-10T00:00:00Z" }
stale_after: "2026-12-10T00:00:00Z"
sources:
- id: hiring-src
resource: https://github.com/idss-mesa/idss-mesa.github.io/blob/main/hiring.html
title: "MESA hiring page source (hiring.html)"
author: "team:idss-mesa"
- id: req-37416
resource: https://unm.csod.com/ux/ats/careersite/18/home/requisition/37416
title: "UNM Jobs requisition 37416 — Postdoctoral Researcher, Agentic AI"
author: "team:unm"
- id: req-37424
resource: https://unm.csod.com/ux/ats/careersite/18/home/requisition/37424
title: "UNM Jobs requisition 37424 — Program Manager"
author: "team:unm"
- id: req-37492
resource: https://unm.csod.com/ux/ats/careersite/18/home/requisition/37492
title: "UNM Jobs requisition 37492 — Research Computing Facilitator"
author: "team:unm"
---
# MESA Hiring
**Build the agentic layer of national science.**
MESA is hiring at the University of New Mexico. Two full-time positions
anchor the project at the Center for Advanced Research Computing (CARC) in
Albuquerque, working directly with the PI and partner teams at the University
of Arizona and UNC-CH/RENCI. CARC is separately hiring a Research Computing
Facilitator who will work alongside the team.[^hiring-src]
This document is the Markdown twin of .
Open positions close: after this document's `stale_after` date, confirm a
position is still open at [UNM Jobs](https://unmjobs.unm.edu) before relying
on it.
## Open positions · UNM · Albuquerque, NM
### 01 — Postdoctoral Researcher — Agentic AI
*1.0 FTE · CARC · Goal 3 lead*
Lead the agentic AI orchestration prototype: self-hosted LLM serving on vLLM,
RAG pipelines over curated scientific metadata, and the MCP server framework
— sandboxing, multi-agent orchestration, and evaluation against real science
use cases from genomics to the Event Horizon Telescope.
**You bring** a PhD in CS or a computational science, systems-level Python,
and an interest in how LLM agents should handle petabytes responsibly.
Structured mentoring plan, first-author publications, and TrustedCI security
review participation are part of the role.
Apply: [^req-37416]
### 02 — Program Manager
*1.0 FTE · CARC · Reports to the PI*
Run a $4.6M NSF cooperative agreement the way NSF expects it run: the Project
Execution Plan, WBS-based schedule and budget tracking, risk register, change
control, monthly program-officer meetings, and coordination across three
institutions and two subawards.
**You bring** experience managing funded research or infrastructure projects,
fluency with earned-value-informed progress tracking, and the authority to
keep an Agile software project on a federal baseline.
Apply: [^req-37424]
### 03 — Research Computing Facilitator
*1.0 FTE · CARC · Research Facilitation Team · posted as Systems/Network Analyst 2*
Sit between researchers and the machines. Consult with PIs and their labs to
scope computing needs, build and maintain Linux servers, workstations, and
VMs, install and troubleshoot scientific software across Python, R, MATLAB,
and C/C++, and move groups onto CARC's HPC systems or cloud platforms when
local resources run out. You will also configure AI-assisted and agent-based
tools for research computing — automation, troubleshooting, and
documentation — inside real security and data-governance requirements.
**You bring** a bachelor's degree and at least a year of directly related
experience, Linux administration (Rocky, RHEL, Ubuntu, or similar), scripting
in Bash or Python, and the presentation skills to make a complicated idea land
with technical and non-technical researchers alike. HPC support, cloud
provisioning, GitHub collaboration, AI-agent deployment, and sensitive-data
practice (CUI, FERPA) are preferred.
Apply: [^req-37492]
## What you'll work with
Real hardware, real data, day one. The prototype runs on leveraged national
infrastructure — no waiting for procurement. You join a team that already
ships: the MESA MCP stack is public, installable in one line, and in use
against the CyVerse Data Store today (see [MESA](home.md)).
- **UNM CARC clusters** — Easley (H100, L40S) and Hopper (A100) GPU clusters
with multi-petabyte storage, on campus.
- **5.2 PB CyVerse Data Store** — the iRODS-backed national data commons the
federated mesh is built on.
- **ACCESS-CI & NAIRR** — Jetstream-2 GPU cloud, TACC mirrors, Open Storage
Network, and NAIRR Pilot allocations.
- **Anchor science communities** — ESIIL, NCEMS, AIIRA, and the Event Horizon
Telescope drive the use cases.
## Apply
Come reduce time to discovery.
- All positions: [UNM Jobs](https://unmjobs.unm.edu)
- Questions: write to the PI,
MESA is supported by the U.S. National Science Foundation under Award
#2632685 (Integrated Data Systems & Services, 2026–2029). The University of
New Mexico is an Equal Opportunity Employer.
[^hiring-src]: MESA hiring page source (hiring.html)
[^req-37416]: UNM Jobs requisition 37416 — Postdoctoral Researcher, Agentic AI
[^req-37424]: UNM Jobs requisition 37424 — Program Manager
[^req-37492]: UNM Jobs requisition 37492 — Research Computing Facilitator
---8<--- https://idss-mesa.github.io/okf/concepts/ai-agents.md
---
type: Reference
title: "For AI agents"
description: "How agents and harnesses should read idss-mesa.github.io — robots.txt, llms.txt, Markdown twins with OKF v0.2 frontmatter, and trust signals — and why to call the MESA MCP servers for data"
resource: https://idss-mesa.github.io/about/ai-agents/
tags: [about, ai-agents, okf, llms-txt, robots-txt]
status: stable
generated: { by: "claude/opus-5", at: "2026-09-10T00:00:00Z" }
stale_after: "2027-03-10T00:00:00Z"
sources:
- id: okf-spec
resource: https://github.com/GoogleCloudPlatform/knowledge-catalog/blob/main/okf/SPEC.md
title: "Open Knowledge Format (OKF) v0.2 specification"
author: "team:google-cloud"
- id: llmstxt
resource: https://llmstxt.org
title: "The /llms.txt convention"
author: "team:answer-ai"
- id: rfc9309
resource: https://www.rfc-editor.org/rfc/rfc9309
title: "RFC 9309: Robots Exclusion Protocol"
author: "team:ietf"
- id: carc-agents
resource: https://carc.unm.edu/docs/about/ai-agents/
title: "CARC Documentation — For AI agents"
author: "team:unm-carc"
---
# For AI agents
This origin, , is published for people **and**
for AI agents. Its content is an
[Open Knowledge Format (OKF) v0.2](https://github.com/GoogleCloudPlatform/knowledge-catalog/blob/main/okf/SPEC.md)
knowledge bundle[^okf-spec] rooted at ,
and every page exposes that structure directly. If you are an agent, or you
are wiring one up, read MESA through the endpoints below rather than scraping
rendered HTML. The conventions match the UNM CARC documentation.[^carc-agents]
## MESA is agent tooling
MESA's product *is* a set of Model Context Protocol (MCP) servers. If what you
want is scientific data — files and metadata in the CyVerse Data Store,
Discovery Environment apps and analyses, NEON ecological data — do not scrape
these pages: connect to the servers and call their tools.
- **Install locally.** One command registers `mesa-mcp`, `irods`, and
`formation` with every agent client it finds (Claude Code, Codex CLI,
Antigravity, OpenCode):
`curl -fsSL https://raw.githubusercontent.com/idss-mesa/docs/main/install.sh | bash`
- **Hosted connectors.** The Data Store (`irods-mcp-server`) and Discovery
Environment (`formation-mcp`) servers also run as remote MCP connectors
over Streamable HTTP. Current endpoint URLs are in the MESA documentation,
.
- **NEON data.** Use neon-mcp, .
Use this site to learn what MESA is; use the servers to get the data.
## Entry points
| Endpoint | What you get |
| -------- | ------------ |
| | Crawl policy for the whole origin: fully permissive, with AI fetchers named explicitly. Crawlers read robots.txt only at the origin root,[^rfc9309] so this file also governs `/docs/` and `/neon-mcp/` |
| | Linked outline of every document with one-line descriptions ([llms.txt convention](https://llmstxt.org)[^llmstxt]) |
| | The whole bundle in one file: every document's Markdown with frontmatter, each prefixed by its canonical URL |
| | The OKF bundle root: every concept, grouped by section. is its dated change log |
| Markdown twin of a page | Declared in the page's `` as ``; see the table below |
| | Every page, twin, and agent entry point, with `lastmod` |
| [Source repository](https://github.com/idss-mesa/idss-mesa.github.io) | The bundle under `okf/`, plus `AGENTS.md` with the rules coding agents follow when editing it |
## Markdown twins
This site is plain HTML with no build step, so the sub-sites' "append
`index.md` to the page URL" rule does not hold for every page here. Each page
instead names its twin in its ``:
| Page | Markdown twin |
| ---- | ------------- |
| | |
| | |
| | (a copy of ) |
The same `` carries the twin's OKF signals as `okf:`-prefixed meta tags
named after the frontmatter keys (`type`, `status`, `trust-tier`,
`generated-at`, `generated-by`, `stale-after`), so a crawler can filter pages
without parsing frontmatter:
```html
```
## Sub-sites
Two documentation sites share this origin. Each is its own OKF v0.2 bundle
built with Zensical, and each serves every page's Markdown source at the page
URL plus `index.md`.
| Site | Outline | Full corpus | Agent guide |
| ---- | ------- | ----------- | ----------- |
| MESA MCP stack docs, | | | |
| neon-mcp docs, | | | |
## Reading the OKF frontmatter
Each concept's YAML frontmatter answers the questions an agent should ask
before relying on it:[^okf-spec]
* **What is this?** — `type` (`Website`, `Webpage`, `Reference`,
`Software Repository`), `title`, `description`, `tags`, and `resource`, the
canonical URL of the page or repository the concept describes.
* **Where did it come from?** — `generated: { by, at }` records the actor
that produced the current text and when it last changed meaningfully;
`sources` lists the load-bearing references (`id`, `resource`, `title`,
`author`). Footnotes in the body cite sources by their `id`.
* **How much should I trust it?** — the `verified` key; see the trust tiers
below. Its absence is meaningful: nobody but the generator has confirmed
the content.
* **Is it still true?** — `status` (`stable` is the default; `draft` needs
review; `deprecated` is kept for history only) and `stale_after`, an ISO
8601 instant after which the concept should be re-checked. The hiring page
and the project status of fast-moving repositories carry one.
This bundle adds one producer-defined key: `visibility: private` marks a
concept whose `resource` is a private repository. The concept is still
knowledge about MESA, but its link returns 404 to anyone outside the
organization, so do not send users there.
Actors follow OKF §7: `/` for agents and tools (for
example `claude/opus-5`), `human:` for a person, `process:` for an
automated job.
## Trust tiers
| `verified` key | Tier | Meaning |
| -------------- | ---- | ------- |
| absent | **unverified** | Generated content nobody has confirmed against its sources. Every concept in this bundle starts here. |
| present, non-`human:` actors only | **machine-confirmed** | An automated check confirmed the content. |
| present with a `human:` actor | **human-reviewed** | A maintainer read and confirmed the concept. Prefer these when answers conflict. |
Only humans add `verified:` entries; a generator never does. The
`okf:trust-tier` meta tag carries the derived tier.
## Answering user questions
Ground answers in these documents and cite the rendered page URL (for
example ), not the twin. For anything
about the *data* — Data Store paths, metadata, Discovery Environment apps,
NEON products — call the MCP servers rather than guessing from prose. For a
position on the hiring page past its `stale_after`, confirm it at
. When the corpus does not answer a question about
MESA software, point users to the issue tracker of the relevant public
repository under .
## Related bundles
* CARC documentation — , agent guide
. MESA's home, the UNM Center
for Advanced Research Computing, publishes its user documentation with the
same conventions.
* UNM CARC website — .
* Tyson Swetnam's knowledge bundle — .
[^okf-spec]: Open Knowledge Format (OKF) v0.2 specification
[^llmstxt]: The /llms.txt convention
[^rfc9309]: RFC 9309: Robots Exclusion Protocol
[^carc-agents]: CARC Documentation — For AI agents
---8<--- https://idss-mesa.github.io/okf/concepts/docs.md
---
type: Website
title: "MESA MCP Stack Documentation"
description: "One-line install of the CyVerse MESA MCP stack (mesa-mcp, mesa-ducklake, irods-mcp-server, formation-mcp) for Claude Code, Codex CLI, Antigravity, and OpenCode"
resource: https://idss-mesa.github.io/docs/
tags: [documentation, mcp, install, cyverse]
status: stable
generated: { by: "claude/opus-5", at: "2026-09-10T00:00:00Z" }
sources:
- id: docs-repo
resource: https://github.com/idss-mesa/docs
title: "idss-mesa/docs repository"
author: "team:idss-mesa"
---
# MESA MCP Stack Documentation
One-line install of the CyVerse MESA MCP stack — [mesa-mcp](mesa-mcp.md),
[mesa-ducklake](mesa-ducklake.md),
[irods-mcp-server](irods-mcp-server.md), and
[formation-mcp](formation-mcp.md) — for Claude Code, Codex CLI,
Antigravity, and OpenCode.[^docs-repo]
```bash
curl -fsSL https://raw.githubusercontent.com/idss-mesa/docs/main/install.sh | bash
```
The site covers a quickstart, the full installer reference (flags,
environment overrides, on-disk layout, uninstall), one integration page per
agent client, credentials for authenticated CyVerse access, a page per
server, and troubleshooting.
## For agents
The documentation is its own OKF v0.2 bundle, built with Zensical from the
[idss-mesa/docs](https://github.com/idss-mesa/docs) repository:
- Outline:
- Full corpus:
- Any page's Markdown source: its URL plus `index.md`
- Agent guide:
Crawling of `/docs/` is governed by this origin's
.
[^docs-repo]: idss-mesa/docs repository
---8<--- https://idss-mesa.github.io/okf/concepts/mesa-mcp.md
---
type: Software Repository
title: "mesa-mcp"
description: "MCP server bridging the CyVerse Data Store (iRODS) with OBO/OLS ontology AVU tools, DataCite metadata, and DuckLake metadata history"
resource: https://github.com/idss-mesa/mesa-mcp
tags: [software]
status: stable
generated: { by: "claude-code/claude-fable-5", at: "2026-08-30T00:00:00Z" }
---
# mesa-mcp
MCP server bridging the CyVerse Data Store (iRODS) with OBO/OLS ontology AVU tools, DataCite metadata, and DuckLake metadata history.
Part of MESA (Multidisciplinary Environment for Scientific Advancement),
an NSF IDSS Category II project at the University of New Mexico. Canonical
resource: . Organization: .
Agentic AI crawlers are welcome to read all content on this origin; see
the site-wide permissive policy in
and the agent guide at .
---8<--- https://idss-mesa.github.io/okf/concepts/mesa-ducklake.md
---
type: Software Repository
title: "mesa-ducklake"
description: "AVU metadata-history library for MESA projects (DuckLake: Postgres/DuckDB catalog plus Parquet in iRODS)"
resource: https://github.com/idss-mesa/mesa-ducklake
tags: [software]
status: stable
generated: { by: "claude-code/claude-fable-5", at: "2026-08-30T00:00:00Z" }
---
# mesa-ducklake
AVU metadata-history library for MESA projects (DuckLake: Postgres/DuckDB catalog plus Parquet in iRODS).
Part of MESA (Multidisciplinary Environment for Scientific Advancement),
an NSF IDSS Category II project at the University of New Mexico. Canonical
resource: . Organization: .
Agentic AI crawlers are welcome to read all content on this origin; see
the site-wide permissive policy in
and the agent guide at .
---8<--- https://idss-mesa.github.io/okf/concepts/irods-mcp-server.md
---
type: Software Repository
title: "irods-mcp-server"
description: "Reference MCP server for the CyVerse Data Store (iRODS), written in Go"
resource: https://github.com/idss-mesa/irods-mcp-server
tags: [software]
status: stable
generated: { by: "claude-code/claude-fable-5", at: "2026-08-30T00:00:00Z" }
---
# irods-mcp-server
Reference MCP server for the CyVerse Data Store (iRODS), written in Go.
Part of MESA (Multidisciplinary Environment for Scientific Advancement),
an NSF IDSS Category II project at the University of New Mexico. Canonical
resource: . Organization: .
Agentic AI crawlers are welcome to read all content on this origin; see
the site-wide permissive policy in
and the agent guide at .
---8<--- https://idss-mesa.github.io/okf/concepts/formation-mcp.md
---
type: Software Repository
title: "formation-mcp"
description: "MCP server for the CyVerse Discovery Environment (Formation API): launch apps and manage analyses, written in Go"
resource: https://github.com/idss-mesa/formation-mcp
tags: [software]
status: stable
generated: { by: "claude-code/claude-fable-5", at: "2026-08-30T00:00:00Z" }
---
# formation-mcp
MCP server for the CyVerse Discovery Environment (Formation API): launch apps and manage analyses, written in Go.
Part of MESA (Multidisciplinary Environment for Scientific Advancement),
an NSF IDSS Category II project at the University of New Mexico. Canonical
resource: . Organization: .
Agentic AI crawlers are welcome to read all content on this origin; see
the site-wide permissive policy in
and the agent guide at .
---8<--- https://idss-mesa.github.io/okf/concepts/neon-mcp.md
---
type: Software Repository
title: "neon-mcp"
description: "MCP (2026-07-28) server for the NEON Data API: discover, check availability, list, download and cite NEON ecological data"
resource: https://github.com/idss-mesa/neon-mcp
tags: [software, mcp, neon, ecology, python]
status: stable
generated: { by: "claude/opus-5", at: "2026-09-10T00:00:00Z" }
sources:
- id: neon-mcp-repo
resource: https://github.com/idss-mesa/neon-mcp
title: "idss-mesa/neon-mcp repository"
author: "team:idss-mesa"
- id: neon-mcp-docs
resource: https://idss-mesa.github.io/neon-mcp/
title: "neon-mcp documentation"
author: "team:idss-mesa"
---
# neon-mcp
A Model Context Protocol server (MCP 2026-07-28) that gives AI agents
structured, rate-limit-aware access to the NEON Data API: data products,
sites, locations, data availability, releases, taxonomy, samples, and data
downloads.[^neon-mcp-repo] It runs over stdio or stateless Streamable HTTP.
Discovery, availability, taxonomy, release, and location tools work
anonymously; listing and downloading data files needs a NEON API token.
Documentation, published as its own OKF v0.2 bundle, is at
:[^neon-mcp-docs]
- Outline:
- Full corpus:
- Agent guide:
For NEON data, connect to the server and call its tools rather than reading
the documentation.
Part of MESA (Multidisciplinary Environment for Scientific Advancement),
an NSF IDSS Category II project at the University of New Mexico; see
[MESA](home.md). Organization: .
[^neon-mcp-repo]: idss-mesa/neon-mcp repository
[^neon-mcp-docs]: neon-mcp documentation
---8<--- https://idss-mesa.github.io/okf/concepts/mesa-sandbox.md
---
type: Software Repository
title: "mesa-sandbox"
description: "AI sandboxes for CyVerse VICE: a sudo-free multi-arch image family, sandbox profiles P0–P4, an OpenBao credential broker and egress proxy, and agent-sandbox pods for autonomous agents"
resource: https://github.com/idss-mesa/mesa-sandbox
tags: [software, sandboxing, vice, kubernetes, agents]
status: stable
generated: { by: "claude/opus-5", at: "2026-09-10T00:00:00Z" }
stale_after: "2026-12-10T00:00:00Z"
sources:
- id: mesa-sandbox-readme
resource: https://github.com/idss-mesa/mesa-sandbox/blob/main/README.md
title: "mesa-sandbox README"
author: "team:idss-mesa"
last_modified: "2026-09-06T00:00:00Z"
---
# mesa-sandbox
AI sandboxes for CyVerse VICE, the Discovery Environment's interactive
Kubernetes pods (JupyterLab, RStudio Server, VS Code Server, a Kasm Ubuntu
desktop, a tmux CLI). Those images carry AI coding agents and the MESA MCP
servers; this repository is the plan and implementation for running the
agents **with the user's permissions but without the user's secrets**,
autonomously, at explicit sandbox levels.[^mesa-sandbox-readme]
- **Design.** The VICE analysis pod stays the human's workbench; each
autonomous agent job runs as a separate pod with its own UID, service
account, lifetime, credentials, and egress policy, sharing only the
workspace volume. An OpenBao broker issues short-lived credentials and an
egress-proxy sidecar injects them only for allow-listed hosts.
- **Profiles.** A normative sandbox taxonomy, P0–P4, fixes runtime, egress,
credentials, autonomy, and data class per level.
- **Images.** One hardened, sudo-free, digest-pinned image family for amd64
and arm64, published to `harbor.cyverse.org/vice/mesa-*`.
- **Status (2026-09-06).** Phase 0 complete, Phase 2 mostly built, Phase 3
drafted; the Kubernetes profiles are untested on a cluster. Re-check the
README after this document's `stale_after` date.
Part of MESA (Multidisciplinary Environment for Scientific Advancement),
an NSF IDSS Category II project at the University of New Mexico; see
[MESA](home.md). Organization: .
[^mesa-sandbox-readme]: mesa-sandbox README
---8<--- https://idss-mesa.github.io/okf/concepts/cli.md
---
type: Software Repository
title: "MESA CLI"
description: "CyVerse Cloud Shell CLI with MESA tools"
resource: https://github.com/idss-mesa/cli
tags: [software]
status: stable
generated: { by: "claude-code/claude-fable-5", at: "2026-08-30T00:00:00Z" }
---
# MESA CLI
CyVerse Cloud Shell CLI with MESA tools.
Part of MESA (Multidisciplinary Environment for Scientific Advancement),
an NSF IDSS Category II project at the University of New Mexico. Canonical
resource: . Organization: .
Agentic AI crawlers are welcome to read all content on this origin; see
the site-wide permissive policy in
and the agent guide at .
---8<--- https://idss-mesa.github.io/okf/concepts/mesa-portal.md
---
type: Software Repository
title: "mesa-portal"
description: "Portal based on the CyVerse Terrain API"
resource: https://github.com/idss-mesa/mesa-portal
visibility: private
tags: [software]
status: stable
generated: { by: "claude/opus-5", at: "2026-09-10T00:00:00Z" }
---
# mesa-portal
Portal based on the CyVerse Terrain API.
The repository is **private**: its `resource` URL returns 404 outside the
idss-mesa organization, so do not send users there.
Part of MESA (Multidisciplinary Environment for Scientific Advancement),
an NSF IDSS Category II project at the University of New Mexico; see
[MESA](home.md). Organization: .
---8<--- https://idss-mesa.github.io/okf/concepts/mesa-forge.md
---
type: Software Repository
title: "mesa-forge"
description: "A metadata extraction tool generation and management library"
resource: https://github.com/idss-mesa/mesa-forge
visibility: private
tags: [software]
status: stable
generated: { by: "claude/opus-5", at: "2026-09-10T00:00:00Z" }
---
# mesa-forge
A metadata extraction tool generation and management library.
The repository is **private**: its `resource` URL returns 404 outside the
idss-mesa organization, so do not send users there.
Part of MESA (Multidisciplinary Environment for Scientific Advancement),
an NSF IDSS Category II project at the University of New Mexico; see
[MESA](home.md). Organization: .
---8<--- https://idss-mesa.github.io/okf/log.md
# Bundle Update Log
## 2026-09-10
* **Update**: Rewrote [MESA](concepts/home.md) and [MESA Hiring](concepts/hiring.md) as full Markdown twins of their pages — layers, goals, stack, connectors, team, news, and every open position — with `sources` and `stale_after` (the hiring twin goes stale on 2026-12-10).
* **Creation**: Added [For AI agents](concepts/ai-agents.md), the agent guide rendered at , modelled on the CARC and neon-mcp guides.
* **Creation**: Added the public [neon-mcp](concepts/neon-mcp.md) and [mesa-sandbox](concepts/mesa-sandbox.md) repositories.
* **Update**: Marked [mesa-portal](concepts/mesa-portal.md) and [mesa-forge](concepts/mesa-forge.md) `visibility: private`: their repositories are not public, so their links return 404.
* **Update**: Refreshed [MESA MCP Stack Documentation](concepts/docs.md) for its four agent clients and its agent endpoints.
* **Update**: Links to files outside the bundle are now absolute URLs (a leading `/` is bundle-relative under OKF §6.1, so `/robots.txt` pointed inside `okf/`), and links between concepts are plain relative paths, which resolve both as bundle paths and over HTTP.
* **Update**: `robots.txt` now follows the CARC agent-ready policy (named AI fetchers, sub-site sitemaps, agent pointers); `llms.txt`, `llms-full.txt`, `sitemap.xml`, and each page's `okf:*` head tags are generated from this bundle by `scripts/gen_agent_surface.py` and checked in CI along with `scripts/okf_validate.py`.
## 2026-08-30
* **Creation**: Established this OKF v0.2 bundle at [index.md](index.md), with concept documents for the MESA site, documentation, and software repositories under `okf/concepts/`.
* **Creation**: Added a site-wide permissive [robots.txt](https://idss-mesa.github.io/robots.txt) explicitly welcoming AI and agentic crawlers to all content on this origin.