ci: shelve v0 e2e workflow — dispatch-only, remove PR trigger#556
Merged
Conversation
The e2e assertion harness has accumulated maintenance debt that causes it to block PRs without actionable signal: - Anthropic API key credit exhaustion → hard failures - Non-deterministic agent-budget regressions (#272) - Auth path reworked twice (#528, #540) without stable baseline Test code and prompts preserved for manual dispatch and reference. A replacement strategy RFQ issue will be filed separately. Co-Authored-By: kevin <kevin@bicameral-ai.com>
Co-Authored-By: kevin <kevin@bicameral-ai.com>
Contributor
Author
🤖 Devin AI EngineerI'll be helping with this pull request! Here's what you should know: ✅ I will automatically:
Note: I can only respond to comments from users who have write access to this repository. ⚙️ Control Options:
|
jinhongkuan
pushed a commit
that referenced
this pull request
Jun 9, 2026
…-intent slash-commands Revives the proven fix from PR #524 (closed only because the v0 e2e CI failed on stale-auth substrate — not the diff; that e2e workflow is now shelved to dispatch-only via #556). Rebased onto current main (all 6 files applied cleanly — nobody touched these since #524 diverged) and the bug re-reproduced on main before fixing: `should_fire_preflight("/qor-plan https://…/issues/1")` returned False. Root cause (Hypothesis 1, confirmed): the UserPromptSubmit classifier in scripts/hooks/preflight_intent.py matched only free-text IMPLEMENTATION_VERBS and had no slash-command awareness. `plan` is not a verb, so `/qor-plan <url>` never fired preflight — planning ran blind to the ledger. Sibling commands worked by coincidence (`/qor-implement`→implement, `/qor-refactor`→refactor). Hypothesis 3 (skill-chain ordering) refuted: UserPromptSubmit fires on raw text before slash-command resolution; the hook ran — its regex was the gate that returned False. Fix: - New IMPL_INTENT_SLASH_COMMANDS frozenset (qor-plan, qor-implement, qor-refactor, qor-debug, qor-remediate, qor-organize, qor-auto-dev-1, qor-auto-dev) — these short-circuit to fire regardless of argument (URL / text / empty). - New layered classify_prompt() → ClassifyResult(fire, prompt_surface_form, slash_command); preflight_reminder.py records prompt_surface_form to telemetry so a future trigger-surface regression is observable. should_fire_preflight() preserved (delegates) for backward compat. - SKILL.md description (Tier-2 caller-LLM gate) updated in lockstep with the Tier-1 hook set, and correctly keeps read-only commands (/qor-status, /qor-help, /qor-audit, /qor-validate) on the SKIP list. Regression guard: test_preflight_intent.py + test_preflight_hook.py now run in test-mcp-regression.yml (they were dormant — in no CI workflow). flow-6 e2e prompt added for reference (the e2e harness is dispatch-only per #556; the unit tests are the durable gate). Closes #402. Supersedes the closed #524. Co-Authored-By: Claude Opus 4.8 (1M context) <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
1 participant
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
Removes the
pull_requesttrigger fromv0-user-flow-e2e.yml, leaving it dispatch-only. The e2e assertion harness has accumulated maintenance debt that blocks PRs without actionable signal:All test code, prompts, and the recording stage are preserved intact — run manually via Actions → "v0 user flow e2e" → Run workflow.
Replacement strategy: #555 (RFQ assigned to @Knapp-Kevin).
Link to Devin session: https://app.devin.ai/sessions/f79f6738778344ae953b97d304e3a4bb