Consolidate canonical AI architecture and ADR decisions for Copiloto development.
npx skills add https://github.com/wagnerra23/oimpresso.com --skill copiloto-arch
Please help me install this Agent Skill. Skill: copiloto-arch Source: https://github.com/wagnerra23/oimpresso.com/tree/main/.claude/skills/jana-arch Command: npx skills add https://github.com/wagnerra23/oimpresso.com --skill copiloto-arch
This Skill removes the need to repeatedly re-read many project ADRs by providing a single, canonical reference for how the Copiloto (Jana) AI architecture is structured and operated.
Use case example: when adding or debugging an Agent in Modules/Copiloto (e.g., ChatCopilotoAgent), use this Skill to confirm which driver stack, memory context angles, telemetry attributes, and multi-tenant rules must be followed.
Ask the AI: “Using the copiloto-arch skill, propose the correct place in Modules/Copiloto to implement this change, and list the required components, tenant rules, metrics updates, and relevant ADR constraints.”
High-intent search queries and answers about installing and using this skill.
Multi-tenant isolation in a Laravel AI architecture is enforced by strictly scoping all queries and memory operations using a business_id. This prevents cross-tenant data leakage while maintaining append-only memory semantics.
Instrumenting OpenTelemetry GenAI events requires applying standardized telemetry conventions to capture AI agent execution attributes. This provides operational visibility into agent performance and contextual tracing within the ERP system.
Configuring Meilisearch for hybrid search involves integrating the search engine with the Laravel AI stack to handle vector and textual queries. This setup must respect tenant boundaries and align with established architectural decision records.
LGPD redaction is required for agent memory in your Laravel AI application to ensure Personally Identifiable Information is masked. This compliance step must be applied before appending data to memory layers or emitting telemetry metrics.
The limitation of append-only memory in AI agent development is that historical context cannot be updated or deleted, requiring strict data hygiene at ingestion. This constraint guarantees audit trails but demands careful PII redaction beforehand.
erphqGovern ERP configuration as code across environments with drift detection and audit trails.
hachejCreate throwaway terminal and UI prototypes to validate software design decisions.
wagnerra23Load canonical oimpresso ERP stack rules and guardrails before starting work.
luanrodriguesOrchestrate a 12-gate multi-tenant development cycle for PHP/Laravel services.
