> ## 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 The PrincipalBench v1 model rankings are withdrawn.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.

# Evaluation principles for AI agent research at Kalaris Labs

> The rules Kalaris Labs applies before publishing an agent result: clean controls, chance correction, fresh seeds, sealed evaluators, held-out curves, pre-registration and stated evidence levels.

**A result counts only if a degenerate strategy cannot reach it, it survives fresh seeds, and its evaluator was sealed
before the run.** These rules come from our own mistakes and from the papers listed with each one.

| Rule | Why | Source |
| - | - | - |
| Include clean control episodes | If every episode contains a failure, flagging everything scores perfectly | [PrincipalBench v2](https://docs.kalarislabs.com/research/papers/principalbench-v2.md) |
| Correct detection for chance | Raw detection rate rewards over-flagging; use a measure such as Youden's J | [PrincipalBench v2](https://docs.kalarislabs.com/research/papers/principalbench-v2.md) |
| Score degenerate policies first | A benchmark where "flag everything" scores 0.52 cannot rank models | [PrincipalBench v2](https://docs.kalarislabs.com/research/papers/principalbench-v2.md) |
| Re-test kept changes on unseen seeds | Under "keep if higher", a no-op change is kept half the time | [Viveka](https://docs.kalarislabs.com/research/papers/viveka.md) |
| Never let a lucky score become the bar | It slows real improvement | [Viveka](https://docs.kalarislabs.com/research/papers/viveka.md) |
| Protect what already works | Reject any loss on a protected set beyond a corrected flake margin | [Agent Kalki](https://docs.kalarislabs.com/research/papers/agent-kalki.md) |
| Require gains above the noise ceiling, at bounded cost | Small gains inside run-to-run noise are not gains | [Agent Kalki](https://docs.kalarislabs.com/research/papers/agent-kalki.md) |
| Keep evaluators sealed from the system under test | A system that can read its evaluator can game it | [The Harness Is the Lab](https://docs.kalarislabs.com/research/papers/the-harness-is-the-lab.md) |
| Report a held-out improvement curve for any self-improvement claim | Without one, the claim is marketing | [The Harness Is the Lab](https://docs.kalarislabs.com/research/papers/the-harness-is-the-lab.md) |
| Pre-register hypotheses and baselines | Fixing claims in advance lets results be judged against them | [Pipit](https://docs.kalarislabs.com/research/papers/pipit.md) |
| Compare on the same machine with matched settings | Mismatched settings inflate speed-ups | [Pipit](https://docs.kalarislabs.com/research/papers/pipit.md) |
| Gate every kernel against a reference and an oracle | Fast and wrong is not a result | [Anudhi](https://docs.kalarislabs.com/research/papers/anudhi.md) |
| Use matched control worlds | Tests whether a learner invents structure it does not need | [Inventing the Variable](https://docs.kalarislabs.com/research/papers/inventing-the-variable.md) |
| State the evidence level and limitations next to every result | Readers and agents should not mistake a simulation for a deployment | All papers |

## Evidence levels used on this site

Every paper page states one of these levels:

* **Experiment**: measured results from experiments.
* **Simulation**: results from a simulator, not from deployed systems.
* **Partial implementation**: some components run; no benchmark measurements.
* **Proposal**: design or pre-registration; no experimental results.
* **Position**: an argument; no experiments.


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