From A Foundry Project

The Foundry workflow creates a no-code rindexer project from contracts that have already been deployed with Foundry scripts. It reads Foundry settings with forge config --json, builds the source with forge build, reads deployment receipts from broadcast/, copies ABI arrays into the generated abis/ directory, and writes a sync file named rindexer-foundry.json.

1. Import a Foundry project

Choose the source form that matches where the Foundry project lives:

from local source

rindexer foundry new ./v4-periphery --output ./uniswap-v4-indexer --name UniswapV4PeripheryImport

When you are already inside a Foundry project, SOURCE is optional and defaults to the current directory, so rindexer foundry new works too. GitHub tree URLs and the generic suffix form <git-url>#<ref>:<subdir> are also supported for nested Foundry projects.

For Git sources, rindexer clones the repository into a temporary directory, checks out the requested ref, and initializes Git submodules when the repository has a .gitmodules file. Private repositories and private submodules use your existing Git credentials. If the source points at a nested directory such as broadcast/, rindexer walks up to the nearest foundry.toml and uses that as the Foundry project root.

2. Deployment requirements

Foundry import is safe generation only. rindexer runs forge build, but it does not deploy contracts and it does not manage private keys.

The Foundry project must build from the source rindexer sees. If a Git import fails because forge build cannot resolve imports, clone the project locally, install dependencies with the repository's setup command, forge install, or forge soldeer install, then import from that local path. For GitHub imports to work from a fresh clone, dependency directories must be committed or declared as Git submodules.

The Foundry project must already contain broadcast deployment receipts. If the import cannot find deployment addresses or receipt block numbers, run your Foundry script locally with --broadcast and retry:

forge script script/Deploy.s.sol --broadcast --rpc-url $RPC_URL

rindexer reads broadcast/*/<chain_id>/run-latest.json first and falls back to the newest run-*.json file for each script and chain. Only CREATE and CREATE2 deployments with events in their ABI are imported. Foundry profile artifacts such as Contract.default.json are matched to their deployed contract name, and test deployments from test/ source paths or contracts ending in Test are skipped to avoid importing fixtures. Deployments are skipped with a reason when their broadcast entry is missing a contract address, receipt block number, matching artifact, or event-bearing ABI. If every deployment is skipped, the command fails and prints the reason summary plus the next action to take.

3. Generated project

The command generates:

rindexer.yaml
rindexer-foundry.json
abis/<Contract>.abi.json
.env
.gitignore
docker-compose.yml
README.md

For example, importing Uniswap v4 periphery:

rindexer foundry new https://github.com/Uniswap/v4-periphery --output ./uniswap-v4-indexer --name UniswapV4PeripheryImport

generates an rindexer.yaml shaped like this:

name: UniswapV4PeripheryImport
description: Generated from a Foundry project.
repository: https://github.com/Uniswap/v4-periphery.git
project_type: no-code
networks:
- name: unichain_sepolia
  chain_id: 1301
  rpc: ${CHAIN_1301_RPC_URL}
- name: base_sepolia
  chain_id: 84532
  rpc: ${CHAIN_84532_RPC_URL}
- name: arbitrum_sepolia
  chain_id: 421614
  rpc: ${CHAIN_421614_RPC_URL}
- name: sepolia
  chain_id: 11155111
  rpc: ${CHAIN_11155111_RPC_URL}
storage:
  postgres:
    enabled: true
native_transfers:
  enabled: false
contracts:
- name: PoolManager
  details:
  - network: unichain_sepolia
    address: 0x00b036b58a818b1bc34d502d3fe730db729e62ac
    start_block: '7092034'
  - network: base_sepolia
    address: 0x05e73354cfdd6745c338b50bcfdfa3aa6fa03408
    start_block: '19088197'
  - network: arbitrum_sepolia
    address: 0xfb3e0c6f74eb1a21cc1da29aec80d2dfe6c9a317
    start_block: '105909222'
  - network: sepolia
    address: 0xe03a1074c86cfedd5c142c4f04f1a1536e203543
    start_block: '7258946'
  abi: ./abis/PoolManager.abi.json
  include_events:
  - name: Initialize
  - name: ModifyLiquidity
  - name: Swap
  - name: Transfer
- name: PositionManager
  details:
  - network: unichain_sepolia
    address: 0xf969aee60879c54baaed9f3ed26147db216fd664
    start_block: '7094939'
  - network: base_sepolia
    address: 0x4b2c77d209d3405f41a037ec6c77f7f5b8e2ca80
    start_block: '20214494'
  - network: arbitrum_sepolia
    address: 0xac631556d3d4019c95769033b5e719dd77124bac
    start_block: '105920123'
  - network: sepolia
    address: 0x429ba70129df741b2ca2a85bc3a2a3328e5c09b4
    start_block: '7259148'
  abi: ./abis/PositionManager.abi.json
  include_events:
  - name: Approval
  - name: ModifyPosition
  - name: Transfer
global:
  health_port: 8080

Each chain ID found in the Foundry broadcast files becomes a rindexer network. Known public chains get readable network names such as ethereum, base, optimism, arbitrum, and polygon; chain 31337 is named anvil; unknown chains fall back to chain_<id>.

The generated rindexer.yaml uses environment-variable placeholders for RPC URLs. Chain 31337 uses ${ANVIL_RPC_URL} and .env defaults it to http://127.0.0.1:8545. Other chains use ${CHAIN_<id>_RPC_URL} placeholders and are intentionally left empty. rindexer will not select, validate, or apply public RPC URLs for you; fill each variable in .env with an endpoint that can serve historic eth_getLogs requests for that chain.

Example generated values:

ANVIL_RPC_URL=http://127.0.0.1:8545
CHAIN_1_RPC_URL=
CHAIN_8453_RPC_URL=

Use an RPC provider with archive access for deployed contracts whose start blocks are far behind the chain tip.

4. Run the generated project

From the generated rindexer project:

rindexer start all

GraphQL is available at http://localhost:3001/graphql when Postgres storage is enabled.

5. Sync later

After adding new broadcasts or changing contract ABIs in the Foundry source, sync the generated rindexer project:

rindexer foundry sync

If you run the command outside the generated project, pass the project path:

rindexer foundry sync --path ./my-indexer

You can preview changes without writing files:

rindexer foundry sync --path ./my-indexer --dry-run

To temporarily sync from a different source, pass it as the optional argument:

rindexer foundry sync https://github.com/org/repo/tree/main/packages/contracts --path ./my-indexer

Sync overwrites managed ABI files and updates the managed contracts in rindexer.yaml. User-owned storage, GraphQL, streams, chat, tables, global settings, and non-Foundry contracts are preserved. Existing .env values are preserved; sync only appends missing RPC keys for newly discovered networks. Stale managed deployments are reported but not deleted.