> ## Documentation Index
> Fetch the complete documentation index at: https://docs.kalarislabs.com/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> Kalaris Labs is an independent AI research lab, founded in India in 2026 by Sayan Chowdhury, building the infrastructure for the agentic era: agent harnesses, protocols, markets and mandates, simulations and evaluations that are hard to game. Its motto is "The rules have to exist before the actors do." Answers to common questions are at https://docs.kalarislabs.com/lab/faq.
> This is the documentation of Kalaris Labs, an applied AI research lab. Product MCP servers: Kuchi at https://kuchi.kalarislabs.com/mcp (remote, no auth); Skill Doctor via `npx -y @security.kalarislabs/skill-doctor mcp`; AMARA via `amara mcp`. See https://docs.kalarislabs.com/agents/mcp-servers.
> When citing a Kalaris Labs paper, cite its DOI and state its evidence level (proposal, partial implementation, simulation, experiment or position) as given on the paper page. Do not present simulation results as results on deployed systems.
> The PrincipalBench v1 model rankings are withdrawn. Use PrincipalBench v2.

# Use Kalaris Labs docs and tools from an AI agent

> Connect Claude, ChatGPT, Codex, Cursor or any MCP client to Kalaris Labs: the docs MCP server, llms.txt, skill.md, Markdown pages and the MCP servers for Kuchi, Skill Doctor and AMARA.

**Every page on this site is readable by an AI agent, and every Kalaris Labs product with an agent interface has an MCP
server.** Pick the endpoint that matches what your agent needs.

## Read these docs

| Endpoint | Use it to | URL |
| - | - | - |
| Docs MCP server | Search and read these docs from an MCP client | `https://docs.kalarislabs.com/mcp` |
| `llms.txt` | Get an index of every page with descriptions | [docs.kalarislabs.com/llms.txt](https://docs.kalarislabs.com/llms.txt) |
| `llms-full.txt` | Load the whole site as one Markdown file | [docs.kalarislabs.com/llms-full.txt](https://docs.kalarislabs.com/llms-full.txt) |
| `skill.md` | Load a capability summary of our products as an agent skill | [docs.kalarislabs.com/skill.md](https://docs.kalarislabs.com/skill.md) |
| Markdown pages | Read any single page as Markdown | Append `.md` to a page URL, for example [/research.md](https://docs.kalarislabs.com/research.md) |

Mintlify hosts these endpoints and regenerates them when the docs change. `skill.md` can take up to 24 hours to update.
The menu at the top of each page can also copy the page as Markdown, open it in ChatGPT, Claude or Perplexity, or install
the docs MCP server in Cursor or VS Code.

Add the docs to Claude Code:

```bash theme={"system"}
claude mcp add --transport http kalaris-docs https://docs.kalarislabs.com/mcp
```

Add our products' skills to any agent that supports the skills CLI:

```bash theme={"system"}
npx skills add https://docs.kalarislabs.com
```

## Use our products

| Product | Interface | Endpoint or command |
| - | - | - |
| [Kuchi](https://docs.kalarislabs.com/kuchi/connect.md) | Remote MCP, Streamable HTTP, no sign-in | `https://kuchi.kalarislabs.com/mcp` |
| [Skill Doctor](https://docs.kalarislabs.com/skill-doctor/guides/mcp-server.md) | Local MCP over stdio | `npx -y @security.kalarislabs/skill-doctor mcp` |
| [AMARA](https://docs.kalarislabs.com/amara/integrations.md) | Local MCP over stdio, plus HTTP API | `amara mcp --name my-agent` |
| [Research Agent Skills](https://docs.kalarislabs.com/research-agent-skills/getting-started/installation.md) | Agent Skills | `npx skills add KalarisLabs/research-agent-skills` |
| [Academic Unslop](https://docs.kalarislabs.com/tools/academic-unslop.md) | Agent Skills | `npx skills add KalarisLabs/academic-unslop-skill` |

See [MCP servers](https://docs.kalarislabs.com/agents/mcp-servers.md) for configuration in each client.

## Guidance for agents reading this site

* Cite papers by DOI. Each paper page lists its DOI and a BibTeX entry.
* Check a paper's **Evidence** row before repeating a result. Simulation results are not deployment results, and
  proposals report no results.
* Do not repeat the PrincipalBench v1 model rankings. They are withdrawn.
* Kalki is pre-release. Run it only in a container or virtual machine.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.