Indexer Handlers

Handler Syntax & Core API

ESM Project

This is an ESM project ("type": "module" in package.json). Top-level await is available. Use import/export syntax, not require.

Modification Workflow

1. After any change to schema.graphql or config.yaml → run pnpm codegen
2. After any change to TypeScript files → run pnpm tsc --noEmit
3. Once compilation succeeds → run pnpm dev to catch runtime errors

Handler Registration

import { indexer } from "envio";

indexer.onEvent(
  { contract: "MyContract", event: "Transfer" },
  async ({ event, context }) => {
    // event.params.<name>  — decoded event parameters
    // event.chainId        — chain ID
    // event.srcAddress     — emitting contract address (checksummed)
    // event.logIndex       — log index within block
    // event.block          — block fields; `number` always, others via `fields`
    // event.transaction    — transaction fields selected via `fields`
  },
);

The first argument is the options object — contract and event names plus optional wildcard / where / fields (see indexer-wildcard, indexer-filters and indexer-transactions skills). The second argument is the handler.

fields lists the block and transaction fields this handler reads. Anything not listed is a type error — see the indexer-transactions skill.

indexer.onEvent(
  { contract: "MyContract", event: "Transfer", fields: { transaction: ["hash"] } },
  async ({ event, context }) => {
    event.transaction.hash; // string
  },
);

Context API

Entity Operations

// Read
const entity = await context.Entity.get(id);              // Entity | undefined
const entity = await context.Entity.getOrThrow(id);       // throws if missing
const entity = await context.Entity.getOrCreate({ id, ...defaults });

// Query by any non-derived field (the index is created on demand)
const list = await context.Entity.getWhere({ fieldName: { _eq: value } });
const list = await context.Entity.getWhere({ fieldName: { _gt: value } });
const list = await context.Entity.getWhere({ fieldName: { _lt: value } });
const list = await context.Entity.getWhere({ fieldName: { _gte: value } });
const list = await context.Entity.getWhere({ fieldName: { _lte: value } });
const list = await context.Entity.getWhere({ fieldName: { _in: [value1, value2] } });
const list = await context.Entity.getWhere({ fieldName: { _gte: min, _lte: max } });
const list = await context.Entity.getWhere({ fieldA: { _eq: a }, fieldB: { _eq: b } });

// Write
context.Entity.set(entity);          // create or update (sync — no await)
context.Entity.deleteUnsafe(id);     // delete (sync — no await)

getWhere operators: _eq, _gt, _lt, _gte, _lte, _in. Multiple fields and operators combine with AND semantics. Any non-derived field is queryable — the indexer creates the matching index the first time it's queried, which pauses indexing while the index builds. Marking a field @index in schema.graphql normally avoids that pause: those indexes are created together when the backfill finishes, before the indexer reports ready. A getWhere on an @index field during backfill still builds it there and then. See indexer-schema for @index syntax.

Context Properties

context.chain.id           // number — current chain ID
context.chain.isRealtime   // boolean — true when ALL chains have caught up to head
context.isPreload      // boolean — true during preload phase
context.log            // { debug, info, warn, error }
context.effect(fn, input)  // external call via Effect API (see indexer-external-calls)

Spread Operator for Updates

Entities from context.Entity.get() are read-only. Always spread:

const entity = await context.Entity.get(id);
if (entity) {
  context.Entity.set({ ...entity, field: newValue });
}

indexer Runtime API

import { indexer } from "envio";

indexer.name;                        // "my-indexer"
indexer.chainIds;                    // [1, 137]
indexer.chains[1].id;                // 1
indexer.chains[1].name;              // "ethereum"
indexer.chains[1].startBlock;        // 0
indexer.chains[1].isRealtime;        // false
indexer.chains[1].MyContract.name;   // "MyContract"
indexer.chains[1].MyContract.addresses; // ["0x..."]
indexer.chains[1].MyContract.abi;    // [...]

Common Pitfalls

Entity IDs — with disable_default_cross_chain: true a row is identified by (id, chainId), so the id only has to be unique within its chain:

const id = `${event.block.number}_${event.logIndex}`;

Without the flag — or on an entity marked @crossChain — the id is the whole key, so prefix it with ${event.chainId}_ to keep chains apart.

Entity relationships — schema uses the entity reference (token0: Token!); handlers use the _id suffix codegen adds (token0_id: token0.id), typed as the referenced entity's id. Never write the bare name (token0) in the handler, and never put _id in the schema.

Optionals — string | undefined, not string | null

Decimal normalization — ALWAYS normalize when adding tokens with different decimals.

Schema & config — see indexer-schema and indexer-configuration skills for full reference.

If something is unclear, use the envio-docs skill to search and read the latest documentation.