# BASE

> Documentation for BASE, the Bittensor subnet (netuid 100): validators run infrastructure; miners compete inside pluggable challenges.

## Docs

- [AGENTS](https://docs.joinbase.ai/AGENTS.md)
- [CONTRIBUTING DOCS](https://docs.joinbase.ai/CONTRIBUTING-DOCS.md)
- [NAV SLUG MAP](https://docs.joinbase.ai/NAV-SLUG-MAP.md)
- [SOURCES](https://docs.joinbase.ai/SOURCES.md)
- [Agent architecture (moved)](https://docs.joinbase.ai/agents/architecture.md): Redirect: agent architecture is under Agent Challenge.
- [baseagent (moved)](https://docs.joinbase.ai/agents/baseagent.md): Redirect: baseagent template docs are under Agent Challenge.
- [Best practices (moved)](https://docs.joinbase.ai/agents/best-practices.md): Redirect: best practices are under Agent Challenge.
- [Agent configuration (moved)](https://docs.joinbase.ai/agents/configuration.md): Redirect: agent configuration is under Agent Challenge.
- [Agent evaluation (moved)](https://docs.joinbase.ai/agents/evaluation.md): Redirect: evaluation is under Agent Challenge.
- [Building agents for Agent Challenge](https://docs.joinbase.ai/agents/overview.md): This tab moved. Agent Challenge miner and builder docs live under Challenges.
- [Agent quickstart (moved)](https://docs.joinbase.ai/agents/quickstart.md): Redirect: Agent Challenge quickstart is under Challenges.
- [Submitting an agent (moved)](https://docs.joinbase.ai/agents/submitting.md): Redirect: agent submit is under Agent Challenge.
- [Tools (moved)](https://docs.joinbase.ai/agents/tools.md): Redirect: agent tools are under Agent Challenge.
- [Broker](https://docs.joinbase.ai/architecture/broker.md): The manager-side broker that dispatches short-lived CPU and GPU evaluation jobs to worker nodes as Swarm replicated-jobs, with a strict GPU contract and untrusted-archive validation.
- [Challenge integration](https://docs.joinbase.ai/architecture/challenge-integration.md): How an isolated challenge plugs into the subnet: the weight contract, per-challenge SQLite state on the /data volume, and the trust boundary with the control plane.
- [Database and registry](https://docs.joinbase.ai/architecture/database.md): The two-tier data model: a shared control-plane PostgreSQL private to the master and per-challenge SQLite on each challenge's /data Swarm volume.
- [Master node](https://docs.joinbase.ai/architecture/master.md): The Docker Swarm manager node that owns the registry, the single public API, the control plane, and final weight computation for the subnet.
- [Architecture overview](https://docs.joinbase.ai/architecture/overview.md): How the BASE subnet is assembled: a single Docker Swarm master node, manually enrolled worker nodes, and the epoch weight flow that ends on-chain.
- [Security model](https://docs.joinbase.ai/architecture/security.md): Trust boundaries, isolation rules, authentication and signing, secret handling, and the Swarm runtime boundary that keep challenges and the control plane apart.
- [Supervisor](https://docs.joinbase.ai/architecture/supervisor.md): The manager-only systemd watchdog that runs the broker-health, timeout-reaper, image-updater, config-sync, and self-update loops for the subnet control plane.
- [Swarm and miner pool](https://docs.joinbase.ai/architecture/swarm.md): The Docker Swarm topology: a single manager node for the control plane and challenge services, plus manually enrolled CPU/GPU worker nodes for short-lived evaluation jobs.
- [Validator network](https://docs.joinbase.ai/architecture/validator-network.md): How validators participate: a submit-only on-chain process that fetches the master's final weight vector and posts it to Bittensor.
- [Weights pipeline](https://docs.joinbase.ai/architecture/weights-pipeline.md): The epoch weight flow: how the master collects raw challenge weights, normalizes emissions, maps hotkeys to Bittensor UIDs, and publishes the vector the submitter posts on-chain.
- [Agent Challenge](https://docs.joinbase.ai/challenges/agent-challenge.md): Hub stub: full Agent Challenge miner documentation lives under Challenges → Agent Challenge.
- [Agent architecture](https://docs.joinbase.ai/challenges/agent-challenge/architecture.md): How an Agent Challenge submission is structured and executed: the entrypoint contract, the baseagent loop, and the isolated runtime.
- [Phala TDX attestation](https://docs.joinbase.ai/challenges/agent-challenge/attestation-phala.md): Intel TDX measurement, dual review/eval images, domain-separated report_data, and trust-but-audit quote accept.
- [baseagent template](https://docs.joinbase.ai/challenges/agent-challenge/baseagent.md): Required Harbor-compatible base implementation for Agent Challenge ZIP submissions.
- [Best practices](https://docs.joinbase.ai/challenges/agent-challenge/best-practices.md): Build Agent Challenge submissions that clear review, self-deploy cleanly, and earn scores.
- [Agent configuration](https://docs.joinbase.ai/challenges/agent-challenge/configuration.md): Configure baseagent-style submissions for Agent Challenge without Base gateway embeds.
- [Evaluation and scoring](https://docs.joinbase.ai/challenges/agent-challenge/evaluation.md): Agent Challenge lifecycle: prepare, deploy, Terminal-Bench cache, score gate, and BASE weights.
- [Agent Challenge](https://docs.joinbase.ai/challenges/agent-challenge/index.md): Miner home for Agent Challenge: package an agent, self-deploy on Phala Intel TDX, and earn weights from attested Terminal-Bench scores.
- [RA-TLS golden key release](https://docs.joinbase.ai/challenges/agent-challenge/key-release.md): How the eval guest obtains the AES-256 golden key over raw TLS 1.3 on :8701, and how client-trust differs from server CA.
- [Agent Challenge quickstart](https://docs.joinbase.ai/challenges/agent-challenge/quickstart.md): Package a baseagent ZIP, submit with a signed hotkey, then self-deploy review and eval on Phala TDX.
- [Submit an agent](https://docs.joinbase.ai/challenges/agent-challenge/submit.md): Package the Agent Challenge ZIP, sign miner requests, and upload through the challenge or BASE proxy.
- [Tools & capabilities](https://docs.joinbase.ai/challenges/agent-challenge/tools.md): The tool surface the baseagent template exposes, and how tools execute inside the Agent Challenge task workspace.
- [Agent Challenge troubleshooting](https://docs.joinbase.ai/challenges/agent-challenge/troubleshooting.md): Common submit, review, eval, and attestation failures for Agent Challenge miners.
- [Agent SWE Challenge](https://docs.joinbase.ai/challenges/agent-swe.md): Miner-facing guide for Agent SWE: real repositories as fail-to-pass software engineering benchmarks.
- [Bounty Challenge](https://docs.joinbase.ai/challenges/bounty-challenge.md): Miner-facing guide for Bounty Challenge: owner-set project briefs, GitHub submissions, human review, emissions.
- [Creating a challenge](https://docs.joinbase.ai/challenges/creating.md): Scaffold a new challenge repository from the BASE template and wire it into the subnet.
- [Data Fabrication Challenge](https://docs.joinbase.ai/challenges/data-fabrication.md): Miner-facing guide for Data Fabrication: submit dataset-generation harnesses, quality over volume.
- [Challenges on BASE](https://docs.joinbase.ai/challenges/overview.md): Challenges are where miners compete. BASE validators run subnet infrastructure; each challenge owns its miner path.
- [PRISM Challenge](https://docs.joinbase.ai/challenges/prism.md): Hub stub: full PRISM miner documentation lives under Challenges → PRISM.
- [PRISM API](https://docs.joinbase.ai/challenges/prism/api.md): The PRISM HTTP API: public challenge routes for submissions and the leaderboard, and the internal bridge, worker, and weights routes.
- [Constraints & limits](https://docs.joinbase.ai/challenges/prism/constraints.md): The PRISM submission limits: the two-script contract, the 150M parameter cap, the AST sandbox, locked data with no network, and the single-node GPU bounds.
- [Providing context](https://docs.joinbase.ai/challenges/prism/context.md): The PrismContext object your architecture and training scripts receive, including geometry, limits, the locked data path, and distributed launch fields.
- [Examples](https://docs.joinbase.ai/challenges/prism/examples.md): A complete, runnable two-script PRISM bundle: the tiny ~1M-parameter example with architecture.py, training.py, and prism.yaml.
- [How PRISM works](https://docs.joinbase.ai/challenges/prism/how-it-works.md): The PRISM service architecture: a FastAPI app, SQLite state, a worker queue, the GPU evaluator, and the weights module.
- [PRISM challenge operators](https://docs.joinbase.ai/challenges/prism/operators.md): Challenge-operator guide for PRISM (not BASE subnet validator mint): local validation, runtime config, data plane, compute budget, LLM gate.
- [PRISM overview](https://docs.joinbase.ai/challenges/prism/overview.md): PRISM is the decentralized neural architecture search challenge that scores how fast a model learns from scratch on locked data.
- [PRISM quickstart](https://docs.joinbase.ai/challenges/prism/quickstart.md): Build a valid two-script PRISM bundle, declare it with prism.yaml, and submit it for forced-init re-execution scoring.
- [Scoring](https://docs.joinbase.ai/challenges/prism/scoring.md): How PRISM scores a submission: prequential bits-per-byte, the held-out delta tie-breaker, the anti-memorization gap, anomaly zeroing, and weights.
- [Submitting to PRISM](https://docs.joinbase.ai/challenges/prism/submit.md): The two-script contract, the prism.yaml manifest, and the three submit modes: full, architecture_only, and training_for_arch.
- [Challenge SDK](https://docs.joinbase.ai/challenges/sdk.md): The shared challenge-side SDK that generated challenges use for evaluation, config, and weights.
- [Challenges explained](https://docs.joinbase.ai/concepts/challenges.md): A challenge is an independent competition plugged into BASE. Miners follow challenge docs; the subnet only coordinates weights.
- [Glossary](https://docs.joinbase.ai/concepts/glossary.md): Definitions of the core BASE and Bittensor terms used throughout these docs.
- [How the subnet works](https://docs.joinbase.ai/concepts/how-it-works.md): From a challenge submission through scoring to on-chain weights on BASE.
- [Incentive mechanism](https://docs.joinbase.ai/concepts/incentive-mechanism.md): How BASE ties rewards to verifiable work through per-challenge scoring and normalized emissions.
- [Installation](https://docs.joinbase.ai/concepts/installation.md): The high-level install paths for BASE, from a local checkout to a validator submitter to a full subnet bring-up.
- [Miners and validators](https://docs.joinbase.ai/concepts/miners-and-validators.md): BASE validators operate the subnet; miners compete inside challenges.
- [What is BASE?](https://docs.joinbase.ai/concepts/overview.md): BASE is the multi-challenge Bittensor subnet on netuid 100: validator infrastructure plus pluggable challenges.
- [Wallets and registration](https://docs.joinbase.ai/concepts/wallets-and-registration.md): The keys you need to participate on BASE, how hotkeys and coldkeys differ, and where registration fits.
- [Weights and emissions](https://docs.joinbase.ai/concepts/weights-and-emissions.md): How BASE turns each challenge's raw scores into a single normalized weight vector for the chain.
- [OG NOTES](https://docs.joinbase.ai/images/OG-NOTES.md)
- [Welcome to BASE](https://docs.joinbase.ai/index.md): BASE is a Bittensor subnet (netuid 100): validators run infrastructure; miners compete inside pluggable challenges.
- [Authentication & signing](https://docs.joinbase.ai/miners/authentication.md): Build the canonical upload message, sign it with your hotkey, and set the four required headers.
- [Choosing a challenge](https://docs.joinbase.ai/miners/choose-a-challenge.md): Compare Agent Challenge, PRISM, and secondary challenges. Each owns its full miner path.
- [Install the miner](https://docs.joinbase.ai/miners/installation.md): Set up the subnet tooling and a Bittensor wallet so you can sign and submit work.
- [Migrating from baseagent](https://docs.joinbase.ai/miners/migration-baseagent.md): Move from standalone baseagent runs to Agent Challenge on BASE: signed submit plus Phala TDX self-deploy.
- [Monitoring & logs](https://docs.joinbase.ai/miners/monitoring.md): Track submission status, watch leaderboards, and check subnet health through the proxy.
- [Miner hub](https://docs.joinbase.ai/miners/overview.md): Register on BASE, choose a challenge, then follow that challenge's miner documentation. There is no undirected Base mining game.
- [Miner quickstart](https://docs.joinbase.ai/miners/quickstart.md): Register on netuid 100, choose a challenge, then follow that challenge's full miner pack.
- [Submitting your work](https://docs.joinbase.ai/miners/submitting.md): Generic subnet proxy upload patterns. Prefer each challenge pack for artifact format and post-submit evaluation.
- [Troubleshooting](https://docs.joinbase.ai/miners/troubleshooting.md): Diagnose and fix upload errors: 401, 409, 413, and 502 responses from the proxy.
- [Wallet & registration](https://docs.joinbase.ai/miners/wallet-registration.md): Create a Bittensor coldkey and hotkey, then register your hotkey on netuid 100.
- [PRISM api (moved)](https://docs.joinbase.ai/prism/api.md): Redirect: PRISM documentation lives under Challenges.
- [PRISM constraints (moved)](https://docs.joinbase.ai/prism/constraints.md): Redirect: PRISM documentation lives under Challenges.
- [PRISM context (moved)](https://docs.joinbase.ai/prism/context.md): Redirect: PRISM documentation lives under Challenges.
- [PRISM examples (moved)](https://docs.joinbase.ai/prism/examples.md): Redirect: PRISM documentation lives under Challenges.
- [PRISM how it works (moved)](https://docs.joinbase.ai/prism/how-it-works.md): Redirect: PRISM documentation lives under Challenges.
- [PRISM operators (moved)](https://docs.joinbase.ai/prism/operators.md): Redirect: PRISM documentation lives under Challenges.
- [PRISM overview (moved)](https://docs.joinbase.ai/prism/overview.md): Redirect: PRISM documentation lives under Challenges.
- [PRISM quickstart (moved)](https://docs.joinbase.ai/prism/quickstart.md): Redirect: PRISM documentation lives under Challenges.
- [PRISM scoring (moved)](https://docs.joinbase.ai/prism/scoring.md): Redirect: PRISM documentation lives under Challenges.
- [PRISM submit (moved)](https://docs.joinbase.ai/prism/submit.md): Redirect: PRISM documentation lives under Challenges.
- [Quickstart (operators and developers)](https://docs.joinbase.ai/quickstart.md): Dry-run BASE master weights with the subnet CLI. Not a miner score path.
- [API overview](https://docs.joinbase.ai/reference/api.md): The public HTTP API exposed by the BASE subnet master, plus its authentication models.
- [Admin API](https://docs.joinbase.ai/reference/api-admin.md): Token-gated and internal endpoints for challenge administration and runtime control on the subnet master.
- [Proxy API](https://docs.joinbase.ai/reference/api-proxy.md): Public proxy routes on the subnet master: the challenge registry, submission status, and challenge passthrough.
- [Weights API](https://docs.joinbase.ai/reference/api-weights.md): The public weights read on the subnet master: the aggregated, chain-ready weight vector for the subnet.
- [CLI overview](https://docs.joinbase.ai/reference/cli.md): The platform command: install, command groups, and the global config option shared by every subcommand.
- [Challenge CLI](https://docs.joinbase.ai/reference/cli-challenge.md): platform challenge commands: scaffold a new challenge repo and register, activate, pull, or restart challenges on the master.
- [DB & registry CLI](https://docs.joinbase.ai/reference/cli-db-registry.md): platform db and platform registry commands: run database migrations, create revisions, and print the live challenge registry.
- [Master CLI](https://docs.joinbase.ai/reference/cli-master.md): platform master commands: run the proxy, broker, and supervisor, refresh and seed challenges, compute weights, and manage Swarm workers.
- [Validator CLI](https://docs.joinbase.ai/reference/cli-validator.md): platform validator run: start a normal validator that evaluates challenges and submits weights on-chain.
- [Configuration examples](https://docs.joinbase.ai/reference/config-examples.md): Annotated master and validator configuration files you can copy and adapt.
- [Configuration](https://docs.joinbase.ai/reference/configuration.md): How Platform loads configuration: YAML files, environment overrides, settings sections, and the production policy.
- [Settings reference](https://docs.joinbase.ai/reference/settings.md): Every Platform configuration key, with its type, default, and source. Grouped by settings section.
- [Upload protocol](https://docs.joinbase.ai/reference/upload-protocol.md): The signed submission protocol miners use to upload to a challenge through the subnet master proxy.
- [Weights schema](https://docs.joinbase.ai/reference/weights-schema.md): Field-level reference for the weight vector models used by the subnet master and challenges.
- [Configuration](https://docs.joinbase.ai/validators/configuration.md): Every validator config field, its default, and where it comes from.
- [Install the validator](https://docs.joinbase.ai/validators/installation.md): Install the platform CLI with the bittensor extra needed to submit weights.
- [Monitoring & logs](https://docs.joinbase.ai/validators/monitoring.md): Read validator log output and confirm weights are being submitted.
- [Validating on BASE](https://docs.joinbase.ai/validators/overview.md): What a validator does on the BASE subnet and how the pieces fit together.
- [Validator quickstart](https://docs.joinbase.ai/validators/quickstart.md): Install, configure, and start a BASE validator from the CLI.
- [Running a validator](https://docs.joinbase.ai/validators/running.md): Start the validator runtime and understand the two loops it runs.
- [The submitter](https://docs.joinbase.ai/validators/submitter.md): The submit-only on-chain weight submitter used on the validator node.
- [Troubleshooting](https://docs.joinbase.ai/validators/troubleshooting.md): Diagnose and fix common validator failures using log lines and config.
- [Weight setting](https://docs.joinbase.ai/validators/weight-setting.md): The validator weight-setting cadence and the on-chain submit path.

## Optional

- [GitHub](https://github.com/BaseIntelligence)