chore(memory): add Phase E + sandbox-gate + runbook-verification lessons#122
Merged
Conversation
Three new entries to .claude/memory.md based on lessons earned in the 2026-05-08 session. - Sandbox local gate (2026-05-08): npm install completes in ~2 min and enables the full five-check gate locally. Saves a CI cycle on first-pass code edits. Build needs SKIP_ENV_VALIDATION=true to bypass scripts/check-env.ts in this sandbox. - Runbook verification rule (2026-05-08): verify every file path / response shape / endpoint behaviour with grep against the actual code before writing it in operator-facing docs. PR #121 first draft had 8 factual errors caught by Copilot review — including pointing operators at a non-existent audit.service.ts and at /api/demo/simulate-whatsapp (which is inbound-only) for outbound WhatsApp testing. - Phase E semantics (2026-05-08): /api/dashboard pendingRecallsCount is "pending recall actions" (sum of dental + vaccination counts), matching the recalls page row count exactly. Future extensions to the recall workspace should preserve the "actions, not horses" semantic — a single horse.count with OR would deduplicate per horse and produce a count that disagrees with /api/recalls totals. Other entries already present from PR #115's memory updates: form-encoded sign-in trap, prod-DB 0-invoices state, 9-template count, Vercel MCP zero-team-scope, 25 UAT cases, lifecycle exercise. https://claude.ai/code/session_01JBMfnqc8QXZRqCG3Tm7hHx
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
Contributor
There was a problem hiding this comment.
Pull request overview
This PR updates the internal session-memory file (.claude/memory.md) to capture three lessons from the 2026-05-08 session, intended to improve future agent workflows (local validation before push, runbook claims verified via grep, and the Phase E “actions not horses” recall-count semantic).
Changes:
- Add a note that local installs in this sandbox are fast enough to run the full local pre-push validation gate.
- Add a note to verify operator runbook claims (paths/endpoints/response shapes) against the codebase before documenting them.
- Add a note clarifying Phase E recall semantics so dashboard and recalls page counts stay aligned.
- Add `npx prisma validate` to the gate list so it matches the "five-check" label and CLAUDE.md's standing rule. - Reword the SKIP_ENV_VALIDATION note: clarify it's a local/sandbox bypass only, never to be set on Vercel project env (which must run check-env normally). https://claude.ai/code/session_01JBMfnqc8QXZRqCG3Tm7hHx
![vercel[bot]](https://avatars.githubusercontent.com/in/8329?s=60&v=4){kind=small}
VADE review nit on PR #122 — standing rule lists prisma validate before test; the session-note line had them swapped. Reordered to lint -> typecheck -> prisma validate -> test -> build for consistency. https://claude.ai/code/session_01JBMfnqc8QXZRqCG3Tm7hHx
Owner
Author
|
@copilot apply changes based on the comments in this thread |
Agent-Logs-Url: https://github.com/RJK134/EquiSmile/sessions/7c4e5837-c56f-4225-8e32-fcabc5b03062 Co-authored-by: RJK134 <167345619+RJK134@users.noreply.github.com>
auto-merge was automatically disabled
May 8, 2026 21:31
Head branch was pushed to by a user without write access
Contributor
Applied and pushed in Warning Firewall rules blocked me from connecting to one or more addresses (expand for details)I tried to connect to the following addresses, but was blocked by firewall rules:
If you need me to access, download, or install something from one of these locations, you can either:
|
Owner
Author
|
@copilot apply changes based on the comments in this thread |
Copilot stopped work on behalf of
RJK134 due to an error
May 8, 2026 21:52
Comment on lines
+95
to
+97
| enables the full local five-check gate (`npm run lint`, | ||
| `npm run typecheck`, `npx prisma validate`, `npm run test`, | ||
| `npm run build`). Run before push to catch component-prop / |
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
Three new entries to
.claude/memory.mdcapturing lessons from the 2026-05-08 session that future Claude sessions need.npm install(~2 min)grepbefore writing operator docsaudit.service.ts, wrong endpoint for outbound WhatsApp) — never write a runbook claim without verifying it firstpendingRecallsCountis "pending actions, not horses"/api/recallsrow count aligned — a singlehorse.countwith OR deduplicates per horse and breaks the alignmentTest plan
.claude/memory.mdand confirms the three new entries are present and dated.Stop-gate
Final merge belongs to Richard or Freddie. Tiny chore PR; safe to merge once CI is green.
https://claude.ai/code/session_01JBMfnqc8QXZRqCG3Tm7hHx
Generated by Claude Code
Note
Low Risk
Low risk documentation-only change: adds session memory notes without modifying runtime code or behavior.
Overview
Adds three new dated entries to
.claude/memory.md: guidance that localnpm installenables running the full pre-push check gate (and thatSKIP_ENV_VALIDATION=trueis needed for Vercel-style builds), a rule togrep/verify paths and endpoint behaviors before writing operator runbooks, and a reminder that Phase E’spendingRecallsCountrepresents pending actions (separate dental/vaccination counts summed) to keep dashboard totals aligned with recall rows.Reviewed by Cursor Bugbot for commit 03ab8be. Configure here.