← ClaudeAtlas

walkie-talkielisted

Audit a completed product feature from the perspective of every affected user and actor, tracing the observable journey through actual code, runtime behavior, data, side effects, and recovery paths. Use after implementing a feature in any product layer, including UI, APIs, background jobs, notifications, authentication, permissions, payments, uploads, synchronization, real-time media, AI or agent actions, integrations, migrations, hardware, mobile, desktop, or web. Apply when validating UX completeness, production readiness, edge cases, lifecycle behavior, concurrency, failure recovery, accessibility, privacy, performance, compatibility, observability, or whether the built implementation matches the intended experience.
varuntej07/tej-stack · ★ 1 · API & Backend · score 74
Install: claude install-skill varuntej07/tej-stack
# Walkie-Talkie > **Update check (run silently before the audit):** run `bash "<this skill's directory>/../../bin/check-update.sh"` and read its single `TEJ_STACK:` status line. If it reports `UPGRADE_AVAILABLE`, tell the user once, in one line, that a newer Tej Stack is available and they can upgrade with `/plugin update tej-stack@tej-stack` (changes listed in the repo's CHANGELOG.md), then continue this task normally. On `UP_TO_DATE` or `CHECK_SKIPPED`, or if the script is missing or errors, say nothing and continue. The check must never block, delay, or fail the task. Audit the feature as experienced, not merely as coded. Reconstruct how every actor enters, understands, uses, waits for, completes, abandons, retries, and recovers from it. Bind every conclusion to implementation or runtime evidence. Default to post-build verification. Inspect and exercise the existing feature before proposing changes. Do not replace evidence with a generic checklist or assume that two implemented endpoints are correctly connected. ## 1. Establish the contract Identify: - primary user intent and definition of success - affected actors, including end users, admins, recipients, external systems, background workers, and future sessions or devices - entry points, triggers, prerequisites, permissions, entitlements, and supported environments - user-visible and invisible outcomes, side effects, and prohibited behavior - shared behavior versus platform-, role-, locale-, account-, or device-sp