Catapult

Catapult is a powerful Ethereum contract deployment and management framework designed to simplify the orchestration of complex contract deployments across multiple blockchain networks. Built with TypeScript and Node.js, it provides a declarative YAML-based approach to defining deployment jobs, templates, and dependencies.

Overview

Catapult addresses the challenge of managing complex contract deployment scenarios where multiple contracts need to be deployed in a specific order, with dependencies between them, across multiple blockchain networks. Instead of writing custom deployment scripts for each scenario, you define your deployment logic declaratively using YAML files.

Key Features

🔄 Declarative Deployment Jobs: Define complex deployment workflows using YAML configuration files
📋 Template System: Create reusable deployment templates that can be shared across projects
🔗 Dependency Management: Automatic resolution of deployment dependencies and execution ordering
🌐 Multi-Network Support: Deploy to multiple blockchain networks simultaneously
⚡ Built-in Actions: Comprehensive set of built-in actions for common deployment tasks
🧮 Value Resolvers: Powerful system for computing values, encoding data, and performing calculations
✅ Skip Conditions: Smart conditional logic to avoid redundant deployments
🔍 Validation & Dry Run: Validate configurations and preview deployment plans without execution
📊 Event System: Rich event system for monitoring deployment progress and debugging
🧾 Multi-platform Verification: Verify on Etherscan v2 and Sourcify (tries all configured platforms by default)

Installation

From npm (recommended)

Available on npm as @0xsequence/catapult.

Global install (provides the catapult CLI on your PATH):

npm install -g @0xsequence/catapult
# or
yarn global add @0xsequence/catapult
# or
pnpm add -g @0xsequence/catapult

Project-local install (use via npx or package.json scripts):

npm install -D @0xsequence/catapult
# or
yarn add -D @0xsequence/catapult
# or
pnpm add -D @0xsequence/catapult

# then run
npx catapult --help

From Source

git clone <repository-url>
cd catapult
npm install
npm run build
npm link

Global Installation (from npm)

npm install -g @0xsequence/catapult

Project Structure

A Catapult project follows this structure:

my-deployment-project/
├── networks.yaml              # Network configurations
├── jobs/                      # Deployment job definitions
│   ├── core-contracts.yaml
│   ├── factory-setup.yaml
│   └── token-deployment.yaml
├── templates/                 # Custom template definitions
│   ├── erc20-factory.yaml
│   └── proxy-deployment.yaml
├── artifacts/                 # Contract build artifacts
│   ├── MyContract.json
│   └── Factory.json
└── output/                    # Generated deployment results

Configuration

Networks Configuration

Create a networks.yaml file in your project root to define target networks:

- name: "Ethereum Mainnet"
  chainId: 1
  rpcUrl: "https://mainnet.infura.io/v3/YOUR_PROJECT_ID"

- name: "Arbitrum One"
  chainId: 42161
  rpcUrl: "https://arb1.arbitrum.io/rpc"

- name: "Polygon"
  chainId: 137
  rpcUrl: "https://polygon-rpc.com"
  supports: ["etherscan_v2"]  # Optional: verification platforms supported
  gasLimit: 500000            # Optional: gas limit for all transactions on this network
  testnet: true               # Optional: mark as test network
  evmVersion: "cancun"        # Optional: network EVM hardfork (e.g., london, paris, shanghai, cancun)

The supports field is optional and specifies which verification platforms are available for the network. Currently supported platforms:

etherscan_v2: Etherscan v2 verification API (supports Ethereum, Polygon, Arbitrum, BSC, etc.)
sourcify: Sourcify verification (no API key required)

If supports is omitted, all built-in platforms are allowed for that network. Etherscan requires an API key to be considered “configured”; Sourcify requires no configuration. The gasLimit field is optional and specifies a fixed gas limit to use for all transactions on this network. If not specified, the system will use ethers.js default gas estimation.

Constants

You can define reusable values in constants files or directly within a job.

Top-level constants are discovered anywhere under your project root by adding YAML files with type: "constants".
Keys must be unique across all constants files; duplicates will fail the load.
Within jobs/templates, reference constants using bare placeholders like {{MY_CONSTANT}}.
Job-level constants override top-level constants when names collide.

Example top-level constants file (can be placed anywhere, e.g., constants.yaml):

type: "constants"

constants:
  address-zero: "0x0000000000000000000000000000000000000000"
  salt-zero: "0x0000000000000000000000000000000000000000000000000000000000000000"
  developer-multisig-01: "0x007a47e6BF40C1e0ed5c01aE42fDC75879140bc4"
  entrypoint-4337-07: "0x0000000071727de22e5e9d8baf0edac6f37da032"

Job-level constants example (defined at the top of a job):

name: "job-with-constants"
version: "1"
constants:
  FEE: "1000"
  ADMIN: "0x0000000000000000000000000000000000000001"
actions:
  - name: "example"
    template: "some-template"
    arguments:
      admin: "{{ADMIN}}"      # resolves to job-level constant
      defaultSalt: "{{salt-zero}}"  # resolves to top-level constant

Tip: Use catapult list constants to see discovered top-level constants and any job-level constants.

RPC URL tokens via environment variables

You can inject secrets (like access tokens) into rpcUrl using placeholders of the form {{RPC_...}}. At load time, any placeholder whose name starts with RPC will be replaced with the value of the corresponding environment variable. Placeholders not starting with RPC are left as-is.

Example networks.yaml:

- name: "MyNet"
  chainId: 999
  rpcUrl: "https://node.url/something/{{RPC_URL_TOKEN}}"

With an environment variable:

export RPC_URL_TOKEN="my-secret-token"

Resulting rpcUrl at runtime:

https://node.url/something/my-secret-token

Notes:

If an {{RPC_*}} placeholder is present and the corresponding environment variable is not set, it now defaults to an empty string. This allows templates like https://node.url/{{RPC_TOKEN}} to collapse gracefully to https://node.url/ without failing the load.
Multiple RPC tokens in one URL are supported, and whitespace inside the token delimiters is ignored (e.g., {{ RPC_TOKEN }}).

Job Definitions

Jobs are the core deployment units. Create YAML files in the jobs/ directory:

# jobs/core-contracts.yaml
name: "core-contracts"
version: "1.0.0"
description: "Deploy core system contracts"

actions:
  - name: "deploy-factory"
    template: "sequence-universal-deployer-2"
    arguments:
      creationCode: "{{Contract(MyFactory).creationCode}}"
      salt: "0"

  - name: "deploy-implementation"
    template: "sequence-universal-deployer-2"
    depends_on: ["deploy-factory"]
    arguments:
      creationCode:
        type: "constructor-encode"
        arguments:
          creationCode: "{{Contract(MyImplementation).creationCode}}"
          types: ["address"]
          values: ["{{deploy-factory.address}}"]
      salt: "0"

Per-job network filters

Jobs run on all selected networks by default. You can restrict or exclude networks for a specific job by chain ID:

name: "token-deployment"
version: "1.0.0"

# Run only on these networks (takes precedence if present)
only_networks: [1, 42161]

# Or, skip these networks (used only if only_networks is not set)
# skip_networks: [137]

actions:
  - name: "deploy"
    template: "erc-2470"
    arguments: { /* ... */ }

Rules:

If only_networks is set and non-empty, the job runs only on those chain IDs.
Else, if skip_networks is set and non-empty, the job is skipped on those chain IDs.
Otherwise, the job runs on all networks selected for the run (via networks.yaml or --network).

Minimum EVM version per job

Jobs can declare a minimum EVM hardfork they require. When a network’s evmVersion is older than the job’s min_evm_version, the job is skipped on that network.

name: "post-shanghai-feature"
version: "1.0.0"
min_evm_version: "shanghai"

actions:
  - name: "deploy"
    template: "erc-2470"
    arguments: { /* ... */ }

Supported identifiers include: frontier, homestead, tangerine, spuriousdragon, byzantium, constantinople, petersburg, istanbul, berlin, london, paris (The Merge), shanghai, cancun, prague.

Deprecating jobs

Mark a job as deprecated to opt it out of normal runs without deleting it:

name: "legacy-seed"
version: "1.2.3"
deprecated: true
actions:
  - name: "noop"
    type: "static"
    arguments: { value: null }

Behavior:

Deprecated jobs are skipped by default when running without specifying job names.
Explicitly targeting a deprecated job on the CLI will run it even without extra flags: catapult run legacy-seed -k $PRIVATE_KEY.
To include all deprecated jobs in a normal run, pass --run-deprecated: catapult run --run-deprecated -k $PRIVATE_KEY.
If a non-deprecated job depends on a deprecated job, that deprecated dependency is ALWAYS included automatically to satisfy dependencies (even without --run-deprecated).

Template Definitions

Templates are reusable deployment patterns. Create them in the templates/ directory:

# templates/proxy-factory.yaml
name: "proxy-factory"
type: "template"

arguments:
  implementation:
    type: "address"
  salt:
    type: "bytes32"

returns:
  address:
    type: "address"

setup:
  - type: "job-completed"
    arguments:
      job: "core-contracts"

actions:
  - type: "send-transaction"
    arguments:
      to: "{{core-contracts.deploy-factory.address}}"
      data:
        type: "abi-encode"
        arguments:
          signature: "createProxy(address,bytes32)"
          values:
            - "{{implementation}}"
            - "{{salt}}"

skip_condition:
  - type: "contract-exists"
    arguments:
      address:
        type: "compute-create2"
        arguments:
          deployerAddress: "{{core-contracts.deploy-factory.address}}"
          salt: "{{salt}}"
          initCode:
            type: "constructor-encode"
            arguments:
              creationCode: "{{Contract(ProxyBytecode).creationCode}}"
              types: ["address"]
              values: ["{{implementation}}"]

outputs:
  address:
    type: "compute-create2"
    arguments:
      deployerAddress: "{{core-contracts.deploy-factory.address}}"
      salt: "{{salt}}"
      initCode:
        type: "constructor-encode"
        arguments:
          creationCode: "{{Contract(ProxyBytecode).creationCode}}"
          types: ["address"]
          values: ["{{implementation}}"]

Notes about template files:

The type: "template" discriminator is optional but recommended for clarity. If provided, it must be exactly template.
Templates are auto-discovered from your project templates/ folder and any templates/ subfolders under jobs/.

Usage

Running Deployments

Deploy all jobs to all configured networks:

catapult run --private-key YOUR_PRIVATE_KEY

Deploy specific jobs:

catapult run core-contracts token-setup --private-key YOUR_PRIVATE_KEY

Deploy multiple jobs using wildcards (matches job names, including nested names like sequence_v3/beta_4):

# Run all jobs whose name starts with "sequence_"
catapult run sequence_* -k $PRIVATE_KEY

# Run all jobs under a namespace/folder-like prefix
catapult run "sequence_v3/*" -k $PRIVATE_KEY

# Combine patterns and exact names; duplicates are de-duplicated
catapult run job1 job? -k $PRIVATE_KEY

Deploy to specific networks:

# Comma-separated, supports chain IDs and network names (name matches include all networks with that name)
catapult run --network 1,42161 --private-key YOUR_PRIVATE_KEY
catapult run --network mainnet --private-key YOUR_PRIVATE_KEY       # all networks named "Mainnet"
catapult run --network mainnet,polygon -k $PRIVATE_KEY core-contracts

Common options (run):

-p, --project <path>: Project root directory (defaults to current directory)
--dotenv <path>: Load environment variables from a custom .env file (run command only)
-n, --network <selectors>: Comma-separated selectors by chain ID or network name
--rpc-url <url>: Run against a single custom RPC; chain ID is auto-detected. If networks.yaml defines that chain, Catapult merges yaml settings (name, supports, gasLimit, testnet, evmVersion, params) while using your RPC URL.
-k, --private-key <key>: EOA private key (or set PRIVATE_KEY)
--etherscan-api-key <key>: Etherscan API key (or set ETHERSCAN_API_KEY)
--fail-early: Stop as soon as any job fails
--ignore-verify-errors: Convert verification errors to warnings and show complete report at end (instead of exiting with error code)
--no-post-check-conditions: Skip post-execution evaluation of skip conditions
--flat-output: Write outputs in a single flat output/ directory (do not mirror jobs/ structure)
--no-summary: Hide the end-of-run summary
--run-deprecated: Allow running jobs marked deprecated: true (otherwise skipped unless explicitly targeted)
--no-std: Do not load built-in standard templates
-v, --verbose (repeatable): Increase logging verbosity (-v, -vv, -vvv)

Examples:

Using a custom RPC (no networks.yaml needed):

catapult run --rpc-url http://127.0.0.1:8545 -k $PRIVATE_KEY

Write outputs flat instead of mirroring jobs/ folders:

catapult run --flat-output -k $PRIVATE_KEY

Run a deprecated job explicitly:
- Without flag (explicit targeting runs it):
```
catapult run legacy-job -k $PRIVATE_KEY
```
- Or include all deprecated jobs in the plan:
```
catapult run --run-deprecated -k $PRIVATE_KEY
```

Validation and Dry Run

Validate your configuration without executing transactions:

catapult dry-run

Validate specific jobs:

catapult dry-run core-contracts --network 1
catapult dry-run core-contracts --network polygon
catapult dry-run core-contracts --network mainnet,42161

Listing Resources

List available jobs:

catapult list jobs

List detected contracts:

catapult list contracts

List available templates:

catapult list templates

List configured networks:

catapult list networks

List only test networks:

catapult list networks --only-testnets

List only non-test networks:

catapult list networks --only-non-testnets

List constants (top-level and per-job):

catapult list constants

Simple outputs for scripting:

# Names only, one per line
catapult list networks --simple

# Chain IDs only, one per line
catapult list networks --simple-chain-ids

Utilities:

# Convert chain ID to network name
catapult utils chain-id-to-name 42161 -p ./my-project

Etherscan helpers:

# Fetch ABI from Etherscan v2
catapult etherscan abi -n 1 -a 0xdAC17F958D2ee523a2206206994597C13D831ec7 --etherscan-api-key $ETHERSCAN_API_KEY
catapult etherscan abi -n mainnet -a 0xdAC17F... --etherscan-api-key $ETHERSCAN_API_KEY

# Fetch source (standard-json or flattened) from Etherscan v2
catapult etherscan source -n 1 -a 0xdAC17F958D2ee523a2206206994597C13D831ec7 --etherscan-api-key $ETHERSCAN_API_KEY
catapult etherscan source -n mainnet -a 0xdAC17F... --etherscan-api-key $ETHERSCAN_API_KEY

Source provenance helpers:

# Verify every source.yaml provenance entry in the project
catapult provenance verify

# Verify only one job's provenance entries
catapult provenance verify my-job

# Verify a job and the jobs it depends on
catapult provenance verify my-job --include-dependencies

# Generate missing build-info JSON files from source.yaml provenance
catapult provenance generate

Built-in Actions

Catapult provides several built-in primitive actions:

`send-transaction`

Send a transaction to the blockchain:

- type: "send-transaction"
  arguments:
    to: "0x742..."
    value: "1000000000000000000"  # 1 ETH in wei
    data: "0x..."
    gasMultiplier: 1.5  # Optional: multiply gas limit by this factor

The gasMultiplier parameter is optional and allows you to tune the gas limit before sending the transaction:

If a network gas limit is configured, it will be multiplied by this factor
If no network gas limit is set, gas will be estimated first, then multiplied by this factor
Must be a positive number (e.g., 1.5 for 50% more gas, 0.8 for 20% less gas)

`send-signed-transaction`

Broadcast a pre-signed transaction:

- type: "send-signed-transaction"
  arguments:
    transaction: "0x..."  # Raw signed transaction

`static`

Sets a static value that can be referenced in subsequent steps. Useful for defining constants or passing data between actions.

- type: "static"
  name: "my-value"
  arguments:
    value: "hello world"

The name field is optional. When provided, the value is stored under name.value in the context. If omitted, the value is computed but not stored. Supports all JSON data types including strings, numbers, booleans, objects, and arrays.

Example with complex data:

- type: "static"
  name: "config"
  arguments:
    value:
      endpoint: "https://api.example.com"
      timeout: 5000
      enabled: true

This makes config.value.endpoint, config.value.timeout, and config.value.enabled available for use in subsequent actions.

`create-contract`

Create a contract by sending its creation bytecode (and optional value):

- type: "create-contract"
  name: "deploy-foo"
  arguments:
    data: "{{Contract(Foo).creationCode}}"
    gasMultiplier: 1.2

`json-request`

Make an HTTP JSON request and use the result downstream:

- type: "json-request"
  name: "get-config"
  arguments:
    url: "https://example.com/config.json"
    method: "GET"

`assert`

Assert an on-chain invariant or value comparison. Evaluates a condition and throws a clear error if it doesn't hold (no broadcast, ever).

The action provides two ways to obtain the ACTUAL value:

to + signature — performs an eth_call (like the call value resolver)
actual — resolves any Value resolver (e.g. read-balance, static, etc.)

Then provide exactly one comparator key (eq, neq, gt, lt, gte, lte) whose value is the EXPECTED result. If the comparison is false, the action throws an error that fails the run.

Example — call a view function and compare:

- type: assert
  name: check-deposit-manager
  arguments:
    to: "{{some-proxy}}"
    signature: "depositManager() returns (address)"
    eq: "{{expected-deposit-manager}}"

Example — resolve a value and compare:

- type: assert
  name: check-balance
  arguments:
    actual: { type: read-balance, arguments: { address: "{{deployer}}" } }
    gte: "1000000000000000000"
    message: "deployer underfunded"

An optional message field is included in the error output for clarity.

Value Resolvers

Catapult includes powerful value resolvers for computing complex values:

`abi-encode`

ABI-encode function call data:

data:
  type: "abi-encode"
  arguments:
    signature: "transfer(address,uint256)"
    values:
      - "0x742..."
      - "1000000000000000000"

`constructor-encode`

Encode constructor parameters with bytecode:

creationCode:
  type: "constructor-encode"
  arguments:
    creationCode: "{{Contract(MyContract).creationCode}}"
    types: ["address", "uint256"]
    values: ["{{factory.address}}", "100"]

`abi-pack`

Pack values per ABI types into bytes:

payload:
  type: "abi-pack"
  arguments:
    types: ["address", "uint256"]
    values: ["{{recipient}}", "{{amount}}"]

`compute-create2`

Compute CREATE2 addresses:

address:
  type: "compute-create2"
  arguments:
    deployerAddress: "{{factory.address}}"
    salt: "{{salt}}"
    initCode: "{{creationCode}}"

`basic-arithmetic`

Perform mathematical operations:

amount:
  type: "basic-arithmetic"
  arguments:
    operation: "add"
    values: ["{{current_balance}}", "1000000000000000000"]

`read-balance`

Read account balance:

balance:
  type: "read-balance"
  arguments:
    address: "{{deployer_address}}"

`call`

Make view/pure function calls:

result:
  type: "call"
  arguments:
    to: "{{contract.address}}"
    signature: "getName()"
    values: []

`get-storage-at`

Read a raw EVM storage slot via eth_getStorageAt:

storageValue:
  type: "get-storage-at"
  arguments:
    address: "{{contract.address}}"
    slot: "0x360894a13ba1a3210667c828492db98dca3e2076cc3735a920a3ca505d382bbc"  # EIP-1967 implementation slot

The slot can be a hex string, a number, or a reference to another value. Returns the 32-byte storage value as a hex string.

`compute-slot`

Compute EVM storage slots for the common Solidity storage layouts. The result is always a 32-byte, 0x-prefixed hex string, so it can be fed directly into get-storage-at or nested as the slot of another compute-slot (e.g. for nested mappings). Select a layout with kind:

mapping — value slot of mapping[key] (keccak256(h(key) . slot)):

slot:
  type: "compute-slot"
  arguments:
    kind: "mapping"
    slot: 0                 # declaration slot of the mapping
    key: "{{owner}}"
    keyType: "address"      # optional, default "uint256"; "string"/"bytes" keys are packed

Nested mappings (balances[a][b]) are expressed by nesting compute-slot in the slot field:

slot:
  type: "compute-slot"
  arguments:
    kind: "mapping"
    key: "{{spender}}"
    keyType: "address"
    slot:
      type: "compute-slot"
      arguments:
        kind: "mapping"
        slot: 1
        key: "{{owner}}"
        keyType: "address"

dynamic-array — element slot of a dynamic array (keccak256(slot) + index * elementSize). The array length lives at slot itself:

slot:
  type: "compute-slot"
  arguments:
    kind: "dynamic-array"
    slot: 3
    index: 4            # optional, default 0
    elementSize: 1      # optional slots-per-element, default 1

struct-field — a struct field or fixed-array element (slot + offset):

slot:
  type: "compute-slot"
  arguments:
    kind: "struct-field"
    slot: "{{structBase}}"
    offset: 2

erc7201 — ERC-7201 namespaced storage root (keccak256(abi.encode(uint256(keccak256(id)) - 1)) & ~0xff):

slot:
  type: "compute-slot"
  arguments:
    kind: "erc7201"
    id: "openzeppelin.storage.Ownable"

eip1967 — well-known EIP-1967 proxy slot (keccak256("eip1967.proxy.<name>") - 1), where name is implementation, admin, or beacon:

implementation:
  type: "get-storage-at"
  arguments:
    address: "{{proxy.address}}"
    slot:
      type: "compute-slot"
      arguments:
        kind: "eip1967"
        name: "implementation"

`verify-contract`

`read-json`

Read a value from a JSON object at a given path:

tokenAddress:
  type: "read-json"
  arguments:
    json: "{{get-config.response}}"
    path: "tokens.usdc.address"

`slice-bytes`

Slice hex strings by byte offsets (supports negative indexes and range syntax):

trimmedPatchData:
  type: "slice-bytes"
  arguments:
    value:
      type: "read-json"
      arguments:
        json: "{{get-guard-v1-signature.response}}"
        path: "txs.data"
    range: ":-1"  # Drop the trailing byte from the payload

You can also provide explicit start and end byte positions (end is exclusive). Negative indexes count from the end of the byte array, so end: -1 trims the last byte and start: -32 keeps the final 32 bytes. range accepts either start:end or the bracket form [start:end].

Verify deployed contracts on block explorers:

- type: "verify-contract"
  arguments:
    address: "{{deploy-factory.address}}"
    contract: "{{Contract(MyContract)}}"  # Reference to the contract to verify
    constructorArguments: "0x000000000000000000000000..."  # Optional hex-encoded args
    platform: "etherscan_v2"  # Optional, defaults to "all" (tries all configured platforms)

Verification Error Handling

By default, verification failures will cause the deployment to exit with an error code. To continue deployment even when verification fails and receive a comprehensive warning report at the end, use the --ignore-verify-errors flag:

catapult run --ignore-verify-errors

This is useful when:

You want to complete all deployments even if some contract verifications fail
Working with networks where verification platforms may be unreliable
Running in CI/CD environments where you prefer warnings over hard failures for verification issues

When enabled, verification errors are converted to warnings and a detailed report is shown at the end of the deployment with all verification failures.

Skip Conditions

Avoid redundant operations with skip conditions:

`contract-exists`

Skip if contract exists at address:

skip_condition:
  - type: "contract-exists"
    arguments:
      address: "{{computed_address}}"

`job-completed`

Skip if another job is completed:

skip_condition:
  - type: "job-completed"
    arguments:
      job: "prerequisite-job"

`skip_if` (pure gate, no post-execution check)

In addition to skip_condition, jobs support a skip_if field for pure gate semantics:

Evaluated ONCE, BEFORE the job runs (the pre-skip decision)
If ANY condition in skip_if is true → skip the whole job (status skipped)
NEVER post-execution-checked (this is the key difference from skip_condition)

Use skip_if for jobs that generate artifacts (e.g., Safe/multisig transaction payloads for human execution out-of-band) and should skip when already in the desired state, without requiring convergence within the run itself.

name: "generate-upgrade-payload"
version: "1.0.0"
skip_if:
  - type: "contract-exists"
    arguments:
      address: "{{computed_upgrade_address}}"
actions:
  - name: "generate"
    type: "static"
    arguments:
      value: "upgrade-payload-data"

Combining skip_condition and skip_if:

If both are present, the job is skipped if ANY condition in either array is true at pre-skip time. Only skip_condition is post-execution-checked.

name: "hybrid-job"
version: "1.0.0"
skip_condition:
  - type: "contract-exists"
    arguments:
      address: "{{deployed_address}}"
skip_if:
  - type: "job-completed"
    arguments:
      job: "setup-job"
actions:
  - name: "deploy"
    template: "erc-2470"
    arguments:
      creationCode: "{{Contract(MyContract).creationCode}}"
      salt: "0"

Standard Templates

Catapult includes several standard templates:

sequence-universal-deployer-2: Deploy contracts using Sequence's Universal Deployer v2
nano-universal-deployer: Deploy contracts using the Nano Universal Deployer
arachnid-deterministic-deployment-proxy: Deploy contracts via Arachnid's CREATE2 proxy at 0x4e59…, with automatic factory bootstrapping
erc-2470 and raw variant: CREATE2 Deployer (singleton factory)
assured-deployment: Helper to ensure a contract is deployed at a specific address
min-balance: Ensure minimum balance for any given address
Raw building blocks: raw-sequence-universal-deployer-2, raw-nano-universal-deployer, raw-erc-2470

Contract Resolution

Catapult automatically discovers and indexes contract artifacts in your project. It supports:

JSON artifacts (Hardhat, Truffle, Foundry)
Nested directory structures
Hash-based contract references
Path-based contract references
Name-based contract references

Reference contracts in your YAML using the new unified Contract() syntax:

creationCode: "{{Contract(path/to/MyContract).creationCode}}"
# or
creationCode: "{{Contract(0x1234...hash).creationCode}}"

Build-info files can carry optional source provenance through a nearby source.yaml sidecar. Deployment runs still use committed artifacts; Catapult does not rebuild from the source repository during catapult run.

jobs/my-stack/build-info/rc-5/
├── stage1.json
└── source.yaml

type: source

build_info:
  "./stage1.json":
    repo: "https://github.com/0xsequence/wallet-contracts-v3"
    ref: "v3.0.0-rc.5"
    commit: "0d9061f229da73edae890e6fdd1fbf753028df6d"
    build: "forge build --build-info"

Catapult can use the same provenance to rebuild and compare build-info files on demand:

# Rebuild each source provenance entry and compare it with the committed file
catapult provenance verify

# Scope to one job, or include that job's dependencies
catapult provenance verify my-job
catapult provenance verify my-job --include-dependencies

# Clone/build from provenance and write missing build-info files
catapult provenance generate
catapult provenance generate my-job --include-dependencies

provenance generate skips existing build-info files by default; pass --force to overwrite them. Both commands clone the configured repo, check out ref or commit, run the build command in that checkout, and look for generated build-info/*.json files. If the build produces more than one build-info file, Catapult selects by matching the committed build-info id, then by filename; if neither is unique, the entry fails with an ambiguity error.

Pinning the build toolchain with `image`

Because provenance compares the entire build-info JSON (including compiler settings such as evmVersion), the rebuild must use the same toolchain that produced the committed file. Add an optional image field to run the build command inside a pinned Docker image instead of on the host:

type: source

build_info:
  "./stage1.json":
    repo: "https://github.com/0xsequence/wallet-contracts-v3"
    commit: "0d9061f229da73edae890e6fdd1fbf753028df6d"
    image: "ghcr.io/foundry-rs/foundry:v1.5.1"
    build: "forge build --build-info"

When image is set, Catapult runs docker run <image> with the checkout bind-mounted at /workspace (also the working directory and $HOME), the build command executed via sh -c, and — on POSIX hosts — the container running as the caller's uid:gid so the generated files are owned by you and the temporary checkout can be cleaned up. This keeps the toolchain pinned per entry (different build-info files can use different images) and leaves the host/runner untouched; it requires Docker to be installed and running. The image field is also supported in per-contract contracts overrides.

If a build-info file needs a per-contract override, key it by fully-qualified contract name:

type: source

build_info:
  "./stage1.json":
    repo: "https://github.com/0xsequence/wallet-contracts-v3"
    commit: "0d9061f229da73edae890e6fdd1fbf753028df6d"
    contracts:
      "src/Stage1Module.sol:Stage1Module":
        ref: "stage1-special"

Output Format

After successful deployment, Catapult generates JSON files in the output/ directory for each job. The output format is optimized to reduce repetition:

Success Grouping

Networks with identical deployment outputs are grouped together:

{
  "jobName": "core-contracts",
  "jobVersion": "1.0.0",
  "lastRun": "2025-01-15T10:30:45.123Z",
  "networks": [
    {
      "status": "success",
      "chainIds": ["1", "42161", "137"],
      "outputs": {
        "deploy-factory.address": "0x742d35Cc6ab8b3c7B3d4B8b3aB4c8f9e9C8e8aB6",
        "deploy-factory.txHash": "0xabc123...",
        "deploy-implementation.address": "0x123abc..."
      }
    }
  ]
}

Error Handling

When deployments fail on specific networks, each failure is recorded separately:

{
  "jobName": "core-contracts", 
  "jobVersion": "1.0.0",
  "lastRun": "2025-01-15T10:30:45.123Z",
  "networks": [
    {
      "status": "success",
      "chainIds": ["1", "42161"],
      "outputs": {
        "deploy-factory.address": "0x742d35Cc6ab8b3c7B3d4B8b3aB4c8f9e9C8e8aB6"
      }
    },
    {
      "status": "error",
      "chainId": "137",
      "error": "Transaction failed: insufficient funds"
    }
  ]
}

This format ensures:

Minimal repetition: Successful deployments with identical outputs across multiple networks are grouped together
Clear error tracking: Individual network failures are clearly documented
Scalability: The format remains readable even with deployments across dozens of networks

Output layout and selection:

By default, output files mirror the structure under jobs/ (e.g., jobs/core/job.yaml -> output/core/job.json). Use --flat-output to write all job JSON files directly under output/.
You can control which action outputs are persisted per job using the output flag on actions:
- output: true to include all outputs for that action
- output: false to exclude outputs for that action
- output: { key1: true, key2: true } to include only specific keys from that action (e.g., txHash, address)

Environment Variables

PRIVATE_KEY: Signer private key (alternative to --private-key)
ETHERSCAN_API_KEY: API key for Etherscan v2 verification (alternative to --etherscan-api-key)

You can load environment variables from a file using --dotenv <path> on the run command (defaults to .env in the current directory when provided).

Development

Prerequisites

Node.js >= 16.0.0
npm or yarn

Setup

# Install dependencies
npm install

# Build the project
npm run build

# Run in development mode
npm run dev

# Watch for changes
npm run watch

Scripts

npm run build - Compile TypeScript to JavaScript
npm run dev - Run the CLI in development mode with ts-node
npm run watch - Watch for changes

Name		Name	Last commit message	Last commit date
Latest commit History 211 Commits
.github/workflows		.github/workflows
contracts		contracts
examples		examples
src		src
test_validation		test_validation
.eslintrc.json		.eslintrc.json
.gitignore		.gitignore
CONCEPT.md		CONCEPT.md
README.md		README.md
eslint.config.js		eslint.config.js
jest.config.js		jest.config.js
package.json		package.json
pnpm-lock.yaml		pnpm-lock.yaml
tsconfig.json		tsconfig.json

Uh oh!

Folders and files

Latest commit

History

Repository files navigation

Catapult

Overview

Key Features

Installation

From npm (recommended)

From Source

Global Installation (from npm)

Project Structure

Configuration

Networks Configuration

Constants

RPC URL tokens via environment variables

Job Definitions

Per-job network filters

Minimum EVM version per job

Deprecating jobs

Template Definitions

Usage

Running Deployments

Validation and Dry Run

Listing Resources

Built-in Actions

send-transaction

send-signed-transaction

static

create-contract

json-request

assert

Value Resolvers

abi-encode

constructor-encode

abi-pack

compute-create2

basic-arithmetic

read-balance

call

get-storage-at

compute-slot

verify-contract

read-json

slice-bytes

Verification Error Handling

Skip Conditions

contract-exists

job-completed

skip_if (pure gate, no post-execution check)

Standard Templates

Contract Resolution

Pinning the build toolchain with image

Output Format

Success Grouping

Error Handling

Environment Variables

Development

Prerequisites

Setup

Scripts

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Uh oh!

Contributors

Uh oh!

Languages

`send-transaction`

`send-signed-transaction`

`static`

`create-contract`

`json-request`

`assert`

`abi-encode`

`constructor-encode`

`abi-pack`

`compute-create2`

`basic-arithmetic`

`read-balance`

`call`

`get-storage-at`

`compute-slot`

`verify-contract`

`read-json`

`slice-bytes`

`contract-exists`

`job-completed`

`skip_if` (pure gate, no post-execution check)

Pinning the build toolchain with `image`

Packages