* fix(workflow): improve error message for runtime API usage in workflow context and fix SKILL.md streaming docs
The workflow stub error now tells users to move calls to a step function or outside the workflow context. Also corrects the SKILL.md streaming section which incorrectly stated getWritable() only works in steps.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* chore: add changeset
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix(skill): address PR feedback on SKILL.md
- Remove `npx workflow start` from debugging section (args can't easily be passed via CLI)
- Mention --json and --help flags for CLI inspection
- Add reminder to check AI docs for DurableAgent details
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix(skill): fix SKILL.md debugging section CLI commands
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* other improvements
* fix(skill): address Copilot PR feedback on streaming examples
- Add missing imports and define all referenced variables
- Wrap writer.releaseLock() in finally blocks for safety
- Remove undefined `data`/`agent`/`messages` references
- Simplify examples to be self-contained and copy-pasteable
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* chore(skill): bump SKILL.md version to 1.1
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* docs: add skill version bump rule to CLAUDE.md
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: John Lindquist <johnlindquist@gmail.com>
- Document fixed versioning for workflow and @workflow/core
- Add genversion pattern for build-time version injection
- Add turbo caching guidance for generated files
- Document executionContext field flexibility
- Document observability data hydration behavior
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
* Add new github workflow
* Enable pull_request trigger for community worlds workflow
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Add community worlds manifest and generation scripts
- Add community-worlds.json manifest as single source of truth
- Add scripts/generate-community-worlds-workflow.mjs to generate CI workflow
- Add scripts/generate-community-worlds-docs.mjs to generate docs section
- Update aggregate-benchmarks.js to load community worlds dynamically
- Add pnpm generate:community-worlds script
- Update docs/deploying/world/index.mdx with community worlds
The manifest-based approach allows:
- E2E tests to be auto-generated from the manifest
- Benchmark aggregation to include community worlds
- Docs to stay in sync with tested worlds
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Fix YAML syntax error - quote strings starting with @
The @ symbol has special meaning in YAML, so package names like
@workflow-worlds/turso need to be quoted.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Fix Redis health-cmd quoting, remove unpublished starter world
- Quote health-cmd when it contains spaces (fixes Docker arg parsing)
- Remove @workflow-worlds/starter as it's not published to npm
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Add benchmarks and summary job for community worlds
- Add build job to share artifacts between benchmark jobs
- Add benchmark jobs for Turso, MongoDB, and Redis worlds
- Update summary job to show both E2E and benchmark status matrix
- Add left border/indent to sidebar child items for visual hierarchy
- Update workflow generator to support benchmark generation
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Reuse build artifacts for E2E tests
E2E jobs now depend on the shared build job and download
artifacts instead of rebuilding packages from scratch.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Add Worlds Ecosystem dashboard to docs
- Create worlds-manifest.json with official and community worlds
- Add aggregate-worlds-data.mjs script for processing E2E and benchmark results
- Create WorldsDashboard, WorldCard, and BenchmarkChart components
- Add /docs/worlds page showing compatibility status and performance
- Include sample data for development
The dashboard shows:
- E2E test progress per world (pass/fail/skip counts)
- Benchmark performance comparison across all worlds
- Filter by official vs community worlds
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Fix CI to output JSON test results and add Jazz world
- Update workflow generator to output JSON test results from vitest
- Upload E2E results as artifacts for parsing in summary job
- Summary job now shows actual pass/fail/skip counts per world
- Add Jazz world to worlds-manifest.json (requires external credentials)
- Add update-worlds-status.yml workflow to auto-update dashboard data
- Update TypeScript types to support null lastRun and metrics
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Refactor community worlds to use reusable workflows
Instead of a generated workflow file, integrate community world testing
directly into tests.yml and benchmarks.yml using reusable workflows.
- Add reusable workflows for E2E tests: e2e-community-world.yml (no services),
e2e-community-world-mongodb.yml, e2e-community-world-redis.yml
- Add reusable workflows for benchmarks: benchmark-community-world.yml,
benchmark-community-world-mongodb.yml, benchmark-community-world-redis.yml
- Update tests.yml to call reusable workflows for Turso, MongoDB, Redis
- Update benchmarks.yml to include community world benchmarks in summary
- Delete generated community-worlds.yml and generator script
This approach:
- Inherits proper Rust/SWC setup from the main workflows
- Keeps all CI in the established patterns
- Makes adding new community worlds straightforward
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Fix benchmark timing file naming for community worlds
Add WORKFLOW_BENCH_BACKEND env var support to bench.bench.ts so community
world benchmarks generate timing files with the correct backend suffix
(e.g., bench-timings-nextjs-turbopack-turso.json instead of -local.json).
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Add @workflow-worlds/starter to community worlds test matrix
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Unify worlds manifest and add dynamic GitHub API fetching
- Merge community-worlds.json into worlds-manifest.json with type field
- Add server-side data fetching from GitHub API for worlds dashboard
- Remove static worlds-status.json, fetch CI artifacts dynamically
- Update all references to use unified manifest format
- Remove obsolete community-worlds.yml and update-worlds-status.yml workflows
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Skip community worlds for non-nextjs-turbopack in benchmark summary
Community worlds only run against nextjs-turbopack, so hide the
"missing" rows for Express and Nitro frameworks in the benchmark
comparison tables.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Fix stream benchmark detection to check for actual TTFB data
The previous check `!== null` incorrectly returned true for undefined,
causing all benchmarks to show TTFB columns. Now explicitly checks
for a number type.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Hide Worlds Ecosystem page from sidebar
The page is still accessible via direct link at /docs/worlds but
won't appear in the navigation until it's been further iterated on.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Fix review comments: trailing newline and division by zero
- Add trailing newline when replacing Community Worlds section in docs
- Fix division by zero in WorldCard benchmark calculation when metrics is empty
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Consolidate community world workflows with service-type parameter
- Create setup-workflow-dev composite action for common setup steps
- Add service-type input to benchmark-community-world.yml and e2e-community-world.yml
- Use conditional job execution (if: inputs.service-type == 'mongodb') to handle different services
- Update benchmarks.yml and tests.yml to pass service-type parameter
- Delete redundant workflow files:
- benchmark-community-world-mongodb.yml
- benchmark-community-world-redis.yml
- e2e-community-world-mongodb.yml
- e2e-community-world-redis.yml
Reduces workflow files from 11 to 7 and eliminates ~500 lines of duplicated YAML.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Generate community world test matrix from worlds-manifest.json
Replace hardcoded community world jobs with dynamic matrix generation using
scripts/create-community-worlds-matrix.mjs. This allows adding/removing
community worlds by editing the manifest instead of multiple workflow files.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Apply suggestion from @vercel[bot]
Co-authored-by: vercel[bot] <35613825+vercel[bot]@users.noreply.github.com>
* Add Samples column and separate local/production benchmarks
- Add Samples column to all benchmark tables showing iteration count
- Separate benchmark results into Local Development and Production sections
- Add explanatory context for each section (localhost vs Vercel deployment)
- Add GitHub action step summaries to e2e community world tests
- Create aggregate-e2e-results.js script for parsing vitest JSON output
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Remove obsolete generate-community-worlds-docs script
The Worlds Ecosystem page now fetches from worlds-manifest.json at runtime,
making this script unnecessary. The npm script also referenced a non-existent
workflow generator script.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Maximize composite action usage and reorganize benchmark output
- Update setup-workflow-dev composite action with optional Rust, install-dependencies, and install-args inputs
- Update tests.yml to use composite action in unit, e2e-vercel-prod, getTestMatrix, e2e-local-*, and getCommunityWorldsMatrix jobs
- Update benchmarks.yml to use composite action in build, benchmark-local, benchmark-postgres, benchmark-vercel, and getCommunityWorldsMatrix jobs
- Reorganize benchmark output to group by benchmark test with local/production tables within each benchmark
- Remove invalid $schema reference from worlds-manifest.json
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Add beads stealth mode stuff (for personal claude memory - will remove stealth if people want)
* Add E2E test results PR comment summary
- Add pr-comment-start job to create/update PR comment when tests start
- Add artifact uploads to all e2e test jobs (vercel-prod, local-dev, local-prod, local-postgres, windows)
- Update e2e-community-world.yml with consistent artifact naming (e2e-community-*)
- Add summary job to aggregate all e2e results and update PR comment
- Extend aggregate-e2e-results.js with --mode aggregate for multi-job PR summary
- Group results by category (Vercel Production, Local Development, etc.)
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Add step summaries to all e2e test jobs
Add "Generate E2E summary" step to each individual e2e job:
- e2e-vercel-prod
- e2e-local-dev
- e2e-local-prod
- e2e-local-postgres
- e2e-windows
Each job now outputs pass/fail/skip counts to GITHUB_STEP_SUMMARY.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Consolidate community world workflows to single job
Replace 3 mutually exclusive jobs (e2e/e2e-mongodb/e2e-redis) with a single
job that starts services via docker run when needed. This eliminates the
skipped job entries that appear in the GitHub Actions UI.
- Use conditional docker run steps instead of services: block
- Add health check loops to wait for service readiness
- Add cleanup step to stop containers
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Fix extractWorldId to handle community world artifact naming
Add handling for `e2e-results-community-{world}` pattern so community
world test results are properly extracted (e.g., `e2e-results-community-turso`
now correctly extracts `turso` instead of `community-turso`).
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
---------
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: vercel[bot] <35613825+vercel[bot]@users.noreply.github.com>
* Proper stacktrace propogation in world
Proper stacktrace propogation in world
* Merge Reconciliation
* Standardize the error type in the world spec
* Deduplicate vercel world utils
* fix undefined type issue
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
* Handle the sleep not completing at the end of the event log
* Handle multiple timers being created in the same suspension
* Better `buildWorkflowSuspensionMessage()`
* New sleep unit tests
---------
Co-authored-by: Nathan Rajlich <n@n8.io>
Nathan Rajlich