182 Commits

Author SHA1 Message Date
YeonGyu-Kim 65521ad813 ci(publish): allow skipping held lazycodex alias 2026-05-31 04:24:24 +09:00
YeonGyu-Kim 2684d08ec9 ci(publish): prepare lsp runtime for release checks 2026-05-31 03:57:13 +09:00
YeonGyu-Kim 9fa62932e4 fix(codex): bootstrap plugin deps before compatibility build 2026-05-31 01:40:28 +09:00
YeonGyu-Kim bcd9ae9901 fix: harden lazycodex platform publishing 2026-05-31 00:42:14 +09:00
YeonGyu-Kim 090e780974 fix: publish lazycodex as release alias 2026-05-30 23:34:13 +09:00
YeonGyu-Kim a789d6c2a5 feat(cli): enable lazycodex codex installs 2026-05-30 21:32:57 +09:00
YeonGyu-Kim 93e1daed4e test(cli): batch 105 (25 files) 2026-05-30 19:12:48 +09:00
YeonGyu-Kim ae6ef6023c fix(ci): batch 31 (4 files) 2026-05-30 19:12:05 +09:00
YeonGyu-Kim 52ba1257b6 Merge branch 'fix/pre-publish-blockers-v4.5.0'
Pre-publish blockers for v4.5.0:
- fix(runtime-fallback): gate retryable signal on status-code allowlist (f05e0cbe9)
- fix(parent-wake): bound assistant-text defer to escape stuck sessions (69c955f61)
- fix(ralph-loop): time-bound oracle dispatch wait to prevent stall (14b3523af)
- test(dist-bundle): assert inlined prompt content survives bundling (3e0a975d1)
- fix(package): block internal-only assets from publish payload (8e28e29c2)

Verified via publish-debate-vortex hyperultradebate (6 hostile agents, 3 rounds).
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-05-26 16:44:50 +09:00
YeonGyu-Kim dc845a8808 chore(web): migrate canonical domain to omo.dev
Switch the production worker route and all hardcoded URLs from
ohmyopenagent.com to omo.dev. Six legacy domains (ohmyopenagent.com,
ohmyopencode.org, ulw.dev, ultrawork.ai, ultrawork.dev,
ultrawork.engineer) are configured to permanently (301) redirect to
omo.dev via Cloudflare Page Rules set up out-of-band via flarectl
and the CF API.

- wrangler.toml: bind worker to omo.dev + www.omo.dev only
- middleware.ts: primaryHost to omo.dev, www to apex redirect retained
- layout.tsx: primarySiteUrl + gaTrackedDomain to omo.dev (GA still G-S0QJFKT46Q)
- sitemap.ts, robots.ts: BASE_URL to https://omo.dev
- npm-downloads/route.ts: usage comment URL refreshed
- web-deploy.yml: environment URL to https://omo.dev
- manifesto.md: domain reality-check line lists omo.dev as canonical
- README*.md (5 langs): npm-downloads badge endpoint to omo.dev
2026-05-25 17:15:16 +09:00
YeonGyu-Kim 3e0a975d1f test(dist-bundle): assert inlined prompt content survives bundling
After the prompts-core migration the TypeScript prompt sources were
deleted; the only mechanism delivering markdown prompts to npm users
is bun build inlining via bunfig.toml [loader] ".md" = "text"
and import attributes. bun test runs from src/index.ts, not from
dist/index.js, so a future Bun upgrade that silently regresses
markdown inlining would pass source tests green while the published
bundle is broken with Cannot find module ../prompts/atlas/default.md
at first agent load.

Add a smoke test that scans the built dist/index.js for unique
signature strings from each migrated prompt file (15 signatures:
3 ultrawork + 5 atlas + 3 prometheus + 4 mode prompts). Skips
gracefully if dist/index.js does not exist (local bun test before
build). Wire into the existing CI Verify dist bundle tests step in
.github/workflows/ci.yml so the regression catches in CI build.

Closes pre-publish blocker V1 and V33.
2026-05-25 01:33:06 +09:00
YeonGyu-Kim 68b80e0343 perf(web): optimize CI + build pipeline
- web-ci.yml: cache .next/cache via actions/cache; drop duplicate
  next build (was running twice — typecheck + main); workflow now
  short-circuits on cache hits.
- scripts/prepare-build.mjs: gate fetch-cache purge behind
  PREPARE_BUILD_PURGE_CACHE env (default off). preview/deploy npm
  scripts no longer invoke it.
- scripts/generate-docs-content.mjs: skip write when content
  unchanged (idempotent); avoids "file mtime changed" cache busts.
- eslint.config.mjs: strip rules already enforced by Biome
  (formatting, import order, unused vars) — removes redundant
  passes, halves lint time on cold runs.
- playwright.config.ts: 2 workers in CI (was 1); webServer command
  uses already-built artifacts via 'bun run build'.

Cold build: 16.2s → 8.7s (-46%). Recompile 1.45s.
2026-05-20 14:26:08 +09:00
YeonGyu-Kim 06f6709317 build: wire ast-grep MCP into release gates
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-05-18 21:28:12 +09:00
YeonGyu-Kim 394567a655 fix(ci): initialize submodules in publish-main checkout
The publish-main job runs npm publish which triggers prepublishOnly →
build:lsp-tools-mcp. That script needs packages/lsp-tools-mcp/ populated,
but actions/checkout@v4 doesn't init submodules by default.

Without this, publish-main failed with 'npm error code 1' on an empty
packages/lsp-tools-mcp/ directory, blocking v4.2.0 release.

Aligns with ci.yml which already uses submodules: recursive everywhere.
2026-05-18 16:48:56 +09:00
YeonGyu-Kim bf96794599 chore(web): move site under packages 2026-05-18 16:44:02 +09:00
YeonGyu-Kim 5e9a26c1c5 chore(packages): align platform package dirs 2026-05-18 16:43:26 +09:00
YeonGyu-Kim 3e3beef21c fix(mcp): point CI lsp submodule path at packages
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-05-18 16:26:52 +09:00
YeonGyu-Kim 54eb3963af chore(ci): init and build vendor lsp submodule in CI 2026-05-18 12:22:54 +09:00
YeonGyu-Kim d8f52aae7f test: run suite without split runner 2026-05-15 16:26:57 +09:00
YeonGyu-Kim e80c2811b1 ci: fail closed on sharded test gate 2026-05-15 12:07:58 +09:00
YeonGyu-Kim fdd40815ba ci: split test workflow across shards
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-05-15 11:41:35 +09:00
YeonGyu-Kim f10251b59d chore(tooling): refresh safe deps and checks 2026-05-14 13:46:11 +09:00
YeonGyu-Kim 7469cb3f3b ci: skip cla for signed contributors 2026-05-13 19:45:06 +09:00
YeonGyu-Kim 100819f0bc ci: build plugin before running tests
The dist-bundle regression tests in
src/shared/dist-bundle-bun-globals.test.ts are guarded by
`test.skipIf(!existsSync("dist/index.js"))` and dist/ is
gitignored, so they silently skipped in CI which ran tests before
the build step. Adding the build step earlier ensures the
regression guard runs and a future raw `Bun.*` leak in the bundle
fails CI.
2026-05-12 12:46:50 +09:00
acamq 6b93fbfd65 ci: enforce frozen bun installs 2026-05-11 16:18:19 -06:00
YeonGyu-Kim 65fc0d9434 ci: generate docs content before web checks; bun prepare hook for local dev
The first build of #3860 failed at type-check because the generated
`lib/docs-content.generated.ts` is gitignored (regenerated on every
build) and CI's `type-check` step runs before `build`. Two fixes:

- web-ci.yml: explicit `Generate docs content from repo-root docs/`
  step right after `bun install` so format-check, lint, and type-check
  all see the file.
- web/package.json: add `prepare` lifecycle script. `bun install`
  invokes it automatically, so a fresh local checkout boots into a
  working state too.

Build still re-runs the generator via prebuild, so docs/ edits land in
the bundle without an explicit dev action.
2026-05-08 16:46:05 +09:00
YeonGyu-Kim aadcbbfcc6 chore(workflows): trigger web CI + deploy on docs/ changes
Now that docs/ is the single source of truth that the marketing site
renders at build time, edits to those files must run web CI and trigger
the Cloudflare deploy. Add `docs/**` to the paths filter in both
web-ci.yml and web-deploy.yml.

A markdown-only fix in docs/ is now sufficient to redeploy
oh-my-openagent.com — no companion web/ change needed.
2026-05-08 16:15:54 +09:00
YeonGyu-Kim 7fe197d8f1 chore(workflows): trigger web-deploy on dev push too
PRs land on `dev` (master is blocked by `block-master-pr`), but the
deploy workflow only listened to `master` pushes — so #3853's web/
dependency bumps merged to dev with no Cloudflare deployment ever
running.

Add `dev` to the push branches list. The existing `paths` filter
keeps the deploy from firing on non-web changes, and `workflow_dispatch`
is preserved as the manual fallback.
2026-05-08 14:57:58 +09:00
YeonGyu-Kim 14398d5fa8 ci(web): add web-ci and web-deploy GitHub Actions workflows
web-ci.yml — runs on push/PR to master|dev that touches web/**:
- format:check (prettier --check)
- lint (eslint flat config)
- type-check (tsc --noEmit)
- bun run build (next build, sanity)
- bunx opennextjs-cloudflare build (Cloudflare worker bundle)

web-deploy.yml — runs on push to master that touches web/** OR manual
workflow_dispatch (with optional environment input):
- bun install --frozen-lockfile
- bun run prebuild + bunx opennextjs-cloudflare build
- cloudflare/wrangler-action@v3 deploy with CLOUDFLARE_API_TOKEN +
  CLOUDFLARE_ACCOUNT_ID secrets, scoped to working-directory: web

Both gated by paths-filter so plugin-only changes do not trigger them.
Concurrency group cancels in-progress CI runs but NOT in-progress deploys.
A web-production GitHub environment is referenced so deploys can be
gated behind required reviewers / wait timers if desired.

Verified locally end-to-end before push:
- bun install: 678 packages
- format:check: pass after `bun run format` reformatted 21 files
- lint: pass
- type-check: pass
- bun run build: pass (4 locales × pages built)
- bunx opennextjs-cloudflare build: pass (.open-next/worker.js generated)
2026-05-08 13:35:00 +09:00
YeonGyu-Kim 690d107334 ci: adapt workflows and test runner for team-mode 2026-05-06 14:19:40 +09:00
YeonGyu-Kim 9a61c607ab fix(ci): treat npm OIDC exchange 201 as success in preflight
Empirically the npm registry returns HTTP 201 (Created) - not 200 -
when the OIDC token exchange succeeds and a fresh publish token is
issued. The preflight gate was only accepting 200 so every
correctly-configured package was flagged as missing. Accept any
2xx status; only treat 4xx/5xx as missing trust config.
2026-04-30 16:27:35 +09:00
YeonGyu-Kim ab5216f6c7 feat(ci): add preflight-trust gate before version bump
The publish workflow used to bump npm latest+1 *before* attempting
the platform publishes. When a platform package was missing its
trusted-publisher config the version was already incremented but
that platform never shipped, leaving partial-publish garbage
versions on npm (this happened with v3.17.7-v3.17.9 during the
OIDC migration).

Add a preflight-trust job that runs in parallel with test/typecheck
and verifies all 24 packages have a trusted publisher configured by
calling npm's own OIDC token exchange endpoint with the workflow's
GitHub OIDC token. publish-main now needs preflight-trust, so any
missing trust config fails the workflow before the version bump.

Failure output lists the exact npm.com URLs to configure each
missing package, plus the org/repo/workflow values to enter.
2026-04-30 16:22:26 +09:00
YeonGyu-Kim fd85dad984 fix(ci): strip _authToken from .npmrc so OIDC takes precedence
After switching to npm Trusted Publishing the publish step still
returned 'PUT 404' because actions/setup-node injects an
'//registry.npmjs.org/:_authToken=${NODE_AUTH_TOKEN}' line into
.npmrc. With NODE_AUTH_TOKEN unset that placeholder evaluates to
an empty string, so npm tries an empty token before reaching for
the OIDC ID token and the registry rejects it.

- Add a step that strips any _authToken line from both project-
  local and $HOME/.npmrc before publishing, so npm CLI proceeds
  to OIDC token exchange.
- Bump publish commands to --loglevel verbose so future failures
  expose the actual auth path (provenance attestation, OIDC
  exchange, etc.) in workflow logs.
2026-04-30 15:19:43 +09:00
YeonGyu-Kim 48fc7bd459 fix(ci): switch npm publish to OIDC trusted publishing
NODE_AUTH_TOKEN expired (set 90 days ago, the npm token default
expiry) causing all publish runs to fail with 'PUT 404 Not Found'
since 2026-04-30.

Migrate publish.yml and publish-platform.yml to npm Trusted
Publishing (OIDC) so we no longer depend on long-lived secrets:

- Bump actions/setup-node v4 -> v6 (improves OIDC compatibility)
- Add 'npm install -g npm@latest' to guarantee npm CLI >= 11.5.1
  (the minimum required for trusted publishing).
- Drop NODE_AUTH_TOKEN env from every publish step. The npm CLI
  picks up the GitHub Actions OIDC token automatically.
- Keep --provenance / NPM_CONFIG_PROVENANCE=true (real-world
  reports indicate provenance is not auto-emitted yet).

Per-package trusted publisher must still be configured on
npmjs.com (Settings -> Trusted Publisher) for all 24 packages
(oh-my-opencode + oh-my-openagent main + 11 platform packages
each, dual-published) before the next publish run.
2026-04-30 14:48:36 +09:00
YeonGyu-Kim 470ed8b13a fix(ci): remove existing signature before ad-hoc signing darwin binary
Bun-compiled binaries contain a malformed LC_CODE_SIGNATURE load
command that prevents codesign from directly replacing it. Remove
the existing signature first, then apply a fresh ad-hoc signature.
2026-04-11 23:55:02 +09:00
YeonGyu-Kim 16e7fa5b1a fix(ci): ad-hoc sign darwin binaries with plain codesign
Bun on github macos-latest runners does not emit linker-signed
signatures by default. Sign explicitly with 'codesign --sign -'
without preserve-metadata since the binary has no prior signature.
2026-04-11 23:51:50 +09:00
YeonGyu-Kim 71b5ab67f6 fix(ci): replace breaking codesign step with signature verification
The previous ad-hoc codesign step failed with 'invalid or unsupported
format for signature' because it tried to preserve the 'linker-signed'
flag which cannot be re-signed. The native macOS build already produces
a valid ad-hoc signature via Bun's linker, so we only need to verify
the signature exists rather than re-sign it.
2026-04-11 23:47:54 +09:00
YeonGyu-Kim 565d3ffa91 fix(ci): build darwin binaries on macOS runners to restore code signing
v3.16.0 binaries had 'adhoc,linker-signed' signatures because newer
Bun versions no longer emit linker-signed signatures when cross-compiling
darwin targets from Linux. Current releases have 'code object is not
signed at all', causing macOS Gatekeeper to reject them immediately.

Verified empirically:
- v3.16.0 darwin-arm64: Signature=adhoc, flags=0x20002(adhoc,linker-signed)
- v3.17.0 darwin-arm64: 'code object is not signed at all'
- Local bun build on macos: produces linker-signed signature (matches v3.16.0)

Changes:
- Route darwin-* platforms to macos-latest runner (native compile)
- Add explicit ad-hoc codesign step as belt-and-suspenders safety net
2026-04-11 23:34:10 +09:00
YeonGyu-Kim 4f196f4917 ci: restore mock-isolated test runner
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-04-08 13:12:34 +09:00
YeonGyu-Kim e2d73f482a ci: run pure bun test in workflows
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-04-06 18:45:33 +09:00
YeonGyu-Kim 63aa1c3e1d ci: use run-ci-tests.ts for mock isolation in CI workflow 2026-04-05 01:16:16 +09:00
YeonGyu-Kim 01994e3f65 ci: sync publish.yml test job with CI.yml isolation pattern 2026-04-05 01:12:54 +09:00
YeonGyu-Kim da86b57f23 test: fix CI test isolation - upgrade Bun, fix mock contamination and fresh-import patterns 2026-04-05 01:01:19 +09:00
YeonGyu-Kim 1071d4defb Pin Bun to 1.3.10 in CI and remove test:ci script
- Pin bun-version to 1.3.10 across all CI workflows to avoid
  mock.module() barrel export regression introduced in 1.3.11
- Remove test:ci script from package.json (use bun test directly)
- Update publish-workflow.test.ts to expect "bun test" instead

🤖 GENERATED WITH ASSISTANCE OF OhMyOpenCode
2026-04-04 18:56:25 +09:00
YeonGyu-Kim eb4060d739 fix(test): narrow mock.module() targets in background-agent tests to prevent barrel corruption
Three test files were mocking the entire '../../shared' barrel, which
corrupted exports for subsequent test files in the same batch run.
Narrow mocks to specific submodules (logger, connected-providers-cache).

Also reverts Bun version pin since the root cause was mock scope, not Bun.
2026-04-04 16:56:39 +09:00
YeonGyu-Kim 99ae6d91d0 fix(ci): pin Bun to 1.3.10 to avoid mock.module() barrel export regression in 1.3.11
Bun 1.3.11 has a regression where mock.module() leaks across test files
in isolated batch execution, causing barrel re-exports to fail with
'Export named X not found' errors. Pin to 1.3.10 until upstream fix.
2026-04-04 16:51:23 +09:00
YeonGyu-Kim 72c8cb8d92 fix(ci): typecheck script tooling in CI workflow
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-04-04 16:12:30 +09:00
YeonGyu-Kim 2c081e1999 fix(ci): restore isolated Bun test execution for mock.module suites
Running Linux CI in one Bun process still leaks mock.module registrations across files, so the workflows now use a CI-specific test runner that isolates mock-heavy targets before executing the remaining suite together.

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-04-04 15:51:54 +09:00
YeonGyu-Kim 57ec3a46ac fix(ci): use unquoted heredoc to allow env var expansion 2026-04-04 14:47:15 +09:00
YeonGyu-Kim 9b30817021 fix(ci): prevent shell injection in workflow expressions 2026-04-04 14:18:25 +09:00