Production schema migrations can fail when existing documents do not match new validators. This skill rehearses the migration on snapshot-seeded preview data before promotion.

Convex incidents can be difficult to explain from large log streams and health events. This skill produces focused findings with evidence and verification links.

Convex onboarding problems can remain hidden inside coding-session transcripts. This skill submits an approved transcript for analysis and summarizes actionable system improvements.

Understanding an unfamiliar Convex application requires connecting schemas, functions, authentication, and integrations. This skill produces a structured, source-grounded architecture map without modifying the application.

Recurring jobs can fail, duplicate work, or expose public functions when schedules and handlers are poorly designed. This skill guides Claude, Codex, and Claude Code to create safe Convex schedules with internal, idempotent, small handlers.

Shot notes are difficult to compare with footage when they live in separate files. This skill combines both into a synchronized video with an animated shot panel.

Firestore Rules can expose private data when access and validation logic are incomplete. This skill helps create structured rules with least-privilege access, schema checks, and security test scenarios.

Manual shot breakdowns are slow and timing estimates are unreliable. This skill measures cuts and motion, then produces structured reviews and interactive reports.

Mobile web applications often feel awkward because browser defaults conflict with touch interaction and changing viewports. This skill applies focused CSS and metadata fixes for a more native mobile experience.

Reusable GTM work needs consistent setup, testing, approvals, and run tracking. This skill creates and maintains workflow projects with cost-aware checks and deployment steps.

GTM context can become inconsistent when organization files, members, and Git history are managed manually. This skill guides creation, updates, validation, migration, and repair with clear approvals.

Manual fit checks produce inconsistent decisions across people and companies. This skill applies saved personas and ICPs to deliver structured, evidence-based qualification.

Teams need consistent buyer personas without mixing organization, ICP, and lead data. This skill creates and maintains scoped Markdown personas in a connected GTM workspace with approval before publishing changes.

Teams can lose ICP decisions across scattered notes and inconsistent workspace files. This skill creates, updates, retires, and checks organization-owned ICP documents with explicit approval before sharing changes.

Complex diffs can hide unnecessary abstractions, dependencies, and dead flexibility. This skill produces concise, location-specific suggestions for deletion or simpler replacements.

SEO changes become difficult to evaluate when queries, pages, dates, or measurements shift. This skill maintains one controlled experiment with comparable observations and scheduled reviews.

Content revisions can weaken meaning or introduce unsupported claims. This skill proposes focused, reviewable edits while preserving voice, provenance, and factual boundaries.

Content and build changes can introduce hidden discovery and structured-data regressions. This skill runs geoptimize and presents deterministic findings with clear evidence and limitations.

Website teams need a careful way to explore emerging discovery files and structured data. This skill previews candidate artifacts and keeps approval with the user before writing changes.

Mailing important PDFs requires correct addresses, service choices, and reliable status checks. This skill submits authorized documents through AgentPostage and reports costs, status, tracking, and receipts.