openwork/.github/workflows/update-models.yml

name: Update Models

on:
  workflow_dispatch:
    inputs:
      base_branch:
        description: Branch to update from and open the pull request against
        required: false
        default: dev
        type: string

permissions:
  contents: write
  pull-requests: write

concurrency:
  group: ${{ github.workflow }}-${{ inputs.base_branch }}
  cancel-in-progress: false

env:
  MODELS_PATH: ee/apps/inference/src/models/base.json
  MODELS_URL: https://models.dev/api.json
  BASE_BRANCH: ${{ inputs.base_branch }}
  BRANCH_NAME: automation/update-models-${{ github.run_id }}-${{ github.run_attempt }}

jobs:
  update-models:
    if: github.repository == 'different-ai/openwork'
    runs-on: blacksmith-4vcpu-ubuntu-2404

    steps:
      - name: Checkout
        uses: actions/checkout@v6
        with:
          ref: ${{ env.BASE_BRANCH }}
          fetch-depth: 1
          token: ${{ github.token }}
          persist-credentials: true

      - name: Setup Node
        uses: actions/setup-node@v6
        with:
          node-version: 24

      - name: Download and validate models
        run: |
          set -euo pipefail

          node --input-type=module <<'NODE'
          import { writeFile } from "node:fs/promises"

          const modelsPath = process.env.MODELS_PATH
          const modelsUrl = process.env.MODELS_URL

          function fail(message) {
            throw new Error(`Refusing to update ${modelsPath}: ${message}`)
          }

          function isRecord(value) {
            return Boolean(value) && typeof value === "object" && !Array.isArray(value)
          }

          function isStringArray(value) {
            return Array.isArray(value) && value.every((item) => typeof item === "string")
          }

          function validateModel(providerId, modelId, model) {
            if (!isRecord(model)) {
              fail(`model ${providerId}.${modelId} was not an object`)
            }

            if (model.id !== modelId) {
              fail(`model ${providerId}.${modelId} was missing a matching id`)
            }

            if (typeof model.name !== "string" || model.name.length === 0) {
              fail(`model ${providerId}.${modelId} was missing a name`)
            }

            if (!isRecord(model.modalities)) {
              fail(`model ${providerId}.${modelId} was missing modalities`)
            }

            if (!isStringArray(model.modalities.input) || model.modalities.input.length === 0) {
              fail(`model ${providerId}.${modelId} was missing input modalities`)
            }

            if (!isStringArray(model.modalities.output) || model.modalities.output.length === 0) {
              fail(`model ${providerId}.${modelId} was missing output modalities`)
            }

            if (!isRecord(model.limit)) {
              fail(`model ${providerId}.${modelId} was missing limits`)
            }
          }

          const response = await fetch(modelsUrl, {
            headers: { accept: "application/json" },
          })

          if (!response.ok) {
            fail(`GET ${modelsUrl} returned ${response.status} ${response.statusText}`)
          }

          const body = await response.text()
          let parsed
          try {
            parsed = JSON.parse(body)
          } catch (error) {
            fail(error instanceof Error ? error.message : "response was not valid JSON")
          }

          if (!isRecord(parsed)) {
            fail("response was not a JSON object")
          }

          const providers = Object.entries(parsed)
          if (providers.length === 0) {
            fail("response did not include any providers")
          }

          for (const [providerId, provider] of providers) {
            if (!isRecord(provider)) {
              fail(`provider ${providerId} was not an object`)
            }

            if (provider.id !== providerId) {
              fail(`provider ${providerId} was missing a matching id`)
            }

            if (typeof provider.name !== "string" || provider.name.length === 0) {
              fail(`provider ${providerId} was missing a name`)
            }

            if (!isRecord(provider.models) || Object.keys(provider.models).length === 0) {
              fail(`provider ${providerId} did not include models`)
            }

            for (const [modelId, model] of Object.entries(provider.models)) {
              validateModel(providerId, modelId, model)
            }
          }

          await writeFile(modelsPath, `${JSON.stringify(parsed)}\n`)
          NODE

      - name: Check for changes
        id: changes
        run: |
          set -euo pipefail

          if git diff --quiet -- "$MODELS_PATH"; then
            echo "changed=false" >> "$GITHUB_OUTPUT"
          else
            echo "changed=true" >> "$GITHUB_OUTPUT"
          fi

      - name: Commit model update
        if: steps.changes.outputs.changed == 'true'
        env:
          COMMIT_SIGNING_KEY: ${{ secrets.COMMIT_SIGNING_KEY }}
        run: |
          set -euo pipefail

          signing_key="$RUNNER_TEMP/commit-signing-key"
          trap 'rm -f "$signing_key"' EXIT
          umask 077
          printf '%s\n' "$COMMIT_SIGNING_KEY" > "$signing_key"
          chmod 600 "$signing_key"

          git config --local user.email "11430621+benjaminshafii@users.noreply.github.com"
          git config --local user.name "benjaminshafii"
          git config --local gpg.format ssh
          git config --local user.signingkey "$signing_key"
          git switch -c "$BRANCH_NAME"
          git add "$MODELS_PATH"
          git commit -S -m "chore: update models snapshot"

      - name: Open pull request
        if: steps.changes.outputs.changed == 'true'
        env:
          GITHUB_TOKEN: ${{ github.token }}
        run: |
          set -euo pipefail

          git push --set-upstream origin "$BRANCH_NAME"
          pr_url="$(gh pr create \
            --base "$BASE_BRANCH" \
            --head "$BRANCH_NAME" \
            --title "chore: update models snapshot" \
            --body "Updates \`$MODELS_PATH\` from \`$MODELS_URL\`.\n\nThe workflow validated the response before replacing the file and wrote the snapshot as minified JSON.")"
          gh pr merge "$pr_url" --auto --squash --delete-branch

      - name: No model changes
        if: steps.changes.outputs.changed == 'false'
        run: echo "No model changes found."