Update campaign workflows and add Codex skill - #1
Open
kenlim-mops wants to merge 18 commits into
Open
kenlim-mops wants to merge 18 commits into
kenlim-mops wants to merge 18 commits into
Conversation
For gated proof-of-concept deployments where an external IdP isn't yet available. Reuses the OIDC signed-session-cookie machinery; a simple domain-restricted, rate-limited email sign-in mints the session and provisions first-time colleagues as low-privilege users. Disabled unless AUTH_PROVIDER=poc; switching to google restores real sign-in with no other change. Access must still be gated at the platform layer. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Runtime migrate() reads drizzle/meta/_journal.json; on Vercel the folder wasn't traced into the function bundle, causing 'Can't find meta/_journal.json' on every DB request. outputFileTracingIncludes ships the migrations with the functions. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Runtime migrate() caused 'Can't find meta/_journal.json' on Vercel (migration folder not traced into serverless functions) and is an anti-pattern anyway. Migrations now run only as an explicit step (db:migrate / poc-provision, gated by RUN_MIGRATIONS_ON_BOOT); the app just connects and queries. PGlite (local/ tests) still auto-migrates. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Adds .claude/skills/utm-builder/ (SKILL.md + reference.md): an Agent Skill that teaches Claude Code and other skill-aware agents to issue governed campaign URLs through /api/v1 instead of hand-crafting utm_* strings. It holds no URL/UTM/ID logic — a thin client of the same registry service as the web app, extension, MCP, and Slack, inheriting canonical IDs, taxonomy, duplicate protection, and audit. Docs updated: new docs/claude-skill.md; README features/clients table/docs index; api.md and mcp.md cross-links; decisions.md ADR 5 client enumeration. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Matches the Codex skill name (.agents/skills/utm-builder-v2) so the two are per-agent siblings of the same capability. Stays in .claude/skills/ (Claude Code's native discovery path); updates all doc references and cross-links the Codex counterpart. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
$utm-builder-v2Codex skill and safe MCP operating workflowValidation
openai.yamlparsed successfullygit diff --checkpassedLint command note
npm run lintcurrently opens Next.js's interactive first-time ESLint configuration because this repository does not contain an ESLint configuration. That prompt was canceled rather than changing lint policy as part of this work. The production build completed its integrated lint/type phase successfully.Merge requirement
The repository requires at least one approving review from a reviewer with write access before merging to
main.