luckyrjain

@luckyrjain source repo

46 skills
0 followers
9 hours ago last updated

▌
Observability Review · luckyrjain bundle
Use when a service's metrics, logs, tracing, dashboards, alerts, SLOs, and correlation IDs need evaluation for coverage and gaps. Keywords: observability review, SLO review, alert coverage, tracing coverage, correlation ID, dashboard review. Not for investigating a live incident (incident-rca) or assessing deployment risk for a specific release (deployment-risk-review).

0
installs
▌
Contract Test Creator · luckyrjain bundle
Generates consumer-driven contract tests (Pact-style) verifying a consumer and provider agree on an interaction shape. Detects the repo's Pact tooling (pact-js, pact-python, Pact JVM, pact-go, Ruby pact) and whether a Pact Broker is configured, then writes either a consumer test (records expectations, produces a pact file) or a provider verification test (replays existing pact files against the real provider) for changed code (diff mode) or an existing coverage gap (backfill mode). Every interaction shape traces to real, observed usage — never a fabricated request/response. Keywords: contract tests, Pact, consumer-driven contract, provider verification, pact broker. Not for a real live integration test against a running dependency (integration-test-creator) or isolated mocked unit tests (unit-test-creator).

0
installs
▌
Incident Triage Agent · luckyrjain bundle
Paging-webhook-triggered composition of incident-rca and squad-map. Page fires → fast triage doc (root cause + owning team) for on-call. Incident resolves → drafted postmortem using incident-rca's own report with squad-map-filled owner columns. No new investigation or ownership logic — delegates entirely to incident-rca and squad-map. Not for interactive, human-typed RCA or ownership requests — those route to incident-rca / squad-map directly. Keywords: PagerDuty, Opsgenie, page fire, on-call triage, postmortem draft, incident-resolved webhook.

0
installs
▌
Loop Task Implementer · luckyrjain bundle
Use when autonomously implementing one or more software tasks through isolated build, evidence-based review, remediation, validation, pull-request, and completion workflows across coding agents. Keywords: implement task/issue, autonomous build+review loop, work through a task queue, take this to PR, builder/reviewer/orchestrator roles. Not for reviewing someone else's already-open MR (pr-review), RCA (incident-rca), or K8s rightsizing (k8s-overprovisioning-datadog).

0
installs
▌
Mysql To Postgres SQL · luckyrjain bundle
Rewrites MySQL-native SQL and datasource/driver config for PostgreSQL during org-wide MySQL scrub. Covers TIMESTAMPDIFF, DATE_FORMAT, DATE_ADD, IFNULL, CAST AS CHAR, CONVERT_TZ, JSON functions, and fulltext search across Java Spring, legacy PHP, Python, and Node.js services. Use when migrating jdbc:mysql/mysql2/psycopg2 connections to PostgreSQL, scrubbing MySQL dialect, auditing native queries, or loading a domain pack for P0/P1 file-level rewrites (timestamps, types, case sensitivity). Keywords: MySQL to PostgreSQL, MySQL-only SQL, MySQL dialect scan, scan-mysql-dialect.sh, mysql2 to pg, SQLAlchemy PG cutover, jdbc:postgresql, domain-pack P0/P1 cooling SQL. Not for a full domain/bounded-context map (domain-comprehension) or MR review (pr-review).

0
installs
▌
Change Impact Analyzer · luckyrjain bundle
Read-only analysis contract for identifying the services, contracts, data, dependencies, owners, tests, and operational surfaces affected by a proposed design or exact PR/MR change. Keywords: change impact, caller/consumer, or affected-service questions; use deployment-risk-review for deployed blast-radius or rollback-risk questions and pr-review for generic correctness review.

0
installs
▌
Deployment Risk Review · luckyrjain bundle
Use when a specific release or change needs a risk assessment before shipping: blast radius, migration risk, rollback complexity, dependency risk, traffic risk, and confidence. Keywords: deployment risk, release risk, blast radius, rollback plan, go/no-go on deployment risk alone. Not for the composed multi-repo release go/no-go sweep (release-readiness-checker), an aggregated ship/no-ship verdict across CI/review/policy/specialists for one change (production-readiness- review), investigating an incident that already happened (incident-triage-agent), or the breaking-change/CVE/API-diff review of a dependency version bump itself (dependency-upgrade-review).

0
installs
▌
Implementation Planner · luckyrjain bundle
Turn approved system design, architecture, change-impact, and triggered specialist evidence into a deterministic single-repository implementation plan with dependency-aware tasks, traceability, execution waves, and loop-task resume compatibility. Keywords: implementation decomposition and DAG planning; use loop-task-implementer to execute the resulting plan.

0
installs
▌
Integration Test Creator · luckyrjain bundle
Generates integration tests that exercise the real seam between a component and one real adjacent dependency (database, queue, cache, internal service) — never a mock of that dependency. Detects the repo's base test runner plus its real-dependency orchestration mechanism (testcontainers, docker-compose, embedded DB) and its integration-test naming/tag convention, writes tests for changed code (diff mode) or an existing coverage gap (backfill mode), runs them against the real dependency, and iterates until green. Keywords: integration tests, testcontainers, docker-compose test, real database test, service seam. Not for isolated mocked tests (unit-test-creator), consumer/provider contract agreements (contract-test-creator), or full browser user journeys (e2e-test-creator).

0
installs
▌
Dependency Upgrade Review · luckyrjain bundle
Use when a framework or library upgrade needs review: breaking changes, CVEs, API differences, transitive dependency impact, and rollout risk. Keywords: dependency upgrade, version bump review, CVE review, breaking change, transitive dependency. Not for a dedicated deep security audit (security-review, which this skill escalates to for exploitable CVEs), the MySQL-to-Postgres migration itself (mysql-to-postgres-sql), or a pre-deploy blast-radius/rollback assessment of the resulting change once the version is picked (deployment-risk-review).

0
installs
▌
Migration Program Manager · luckyrjain bundle
Org-wide rollup of mysql-to-postgres-sql's MIGRATION_STATUS.yaml across many workspaces, joined to squad-map's SQUAD_MAP.md for ownership, ranked by staleness and blocked-gate count per squad. Keywords: migration program, migration status across repos, stalled migrations, migration rollup, org-wide migration tracking. Not for one workspace's own migration status (mysql-to-postgres-sql directly) or squad/repo ownership lookups (squad-map).

0
installs
▌
Release Readiness Checker · luckyrjain bundle
Release go/no-go report composing pr-review (MRs merged since last release, never posts), k8s- overprovisioning-datadog (per-service rightsizing verdict), and incident-rca (per-service open-incident signal, Phase 1 only). Keywords: release readiness, is this release ready, ship checklist, release go/no-go, pre-release check. Not for reviewing one specific MR (pr-review), one service's rightsizing (k8s-overprovisioning-datadog), or a full root-cause investigation (incident-rca).

0
installs
▌
Production Readiness Review · luckyrjain bundle
Read-only orchestrator answering "is this PR/MR/release-candidate production ready?" by gathering trusted evidence (CI, code review, build provenance, SCM policy, change-impact, deployment-risk) and dispatching the applicable specialist reviews (security, observability, resilience, API design, database, performance, capacity, dependency-upgrade), then aggregating everything into one fail-closed verdict. Keywords: production ready, ready to release, ship this PR, go/no-go for one change. Not for generic code review (pr-review), a multi-repo/multi-service release sweep (release-readiness-checker), or a standalone change-impact/blast-radius/deployment-risk question (change-impact-analyzer, deployment-risk-review).

0
installs
▌
Codebase Architecture Review · luckyrjain bundle
Use when an existing codebase needs a bounded, evidence-backed review of architecture friction and potential refactoring candidates. Keywords: codebase architecture review, architecture friction, refactoring opportunities, change locality, coupling, cohesion. Not for a proposed architecture (architecture-review), one module's design (module-design), or implementation work.

0
installs
▌
K8S Overprovisioning Datadog · luckyrjain bundle
Use when the user asks whether a Kubernetes deployment or service is overprovisioned, right-sized, underprovisioned, or ready for resource optimization using Kubernetes or observability MCP data. Keywords: Kubernetes MCP, Datadog, overprovisioned, right-size, rightsizing, CPU/memory requests, HPA, replicas, throttling, OOM, Kafka consumer lag, cost/waste, namespace waste ranking. Not for root-cause/outage investigation (incident-rca), MR review (pr-review), or applying manifest changes (recommendations only).

0
installs
▌
Engineering Decision Discovery · luckyrjain bundle
Use when engineering decisions remain unresolved and need an interactive, evidence-backed challenge before design or implementation. Keywords: grill me, challenge my plan, stress-test this decision, question my assumptions, help me decide, what decisions are missing, interrogate this architecture. Not for reconstructing current domain behavior, reviewing a proposed architecture, or implementing an already-settled task.

0
installs
▌
Cost Optimization Sprint Planner · luckyrjain bundle
Org-wide cost/waste sweep composing k8s-overprovisioning-datadog once per in-scope deployment, ranked by monthly_savings_total descending and grouped by squad. Optional namespace-ranking pre-filter before running full per-deployment assessments. Keywords: cost optimization sprint, org-wide waste ranking, where's the money, rightsizing sprint planning, cost savings backlog. Not for one deployment's own rightsizing question (k8s-overprovisioning-datadog directly) or squad/repo ownership lookups (squad-map).

0
installs
▌
Pr Review · luckyrjain bundle
GitHub pull-request and GitLab merge-request review by URL, number, or current branch. Phased workflow: context → findings pipeline → executive summary → optional comment post. Keywords: /pr-review, review PR/MR/pull request/merge request, re-review, post-merge audit, list open reviews, review and post. Supports GitHub.com, GitHub Enterprise Server, GitLab.com, and self-hosted GitLab. Not for local-only diffs, RCA, or K8s rightsizing. Full phrase table: examples.md.

0
installs
▌
Squad Map · luckyrjain bundle
Maps repos to org squads (GitLab group hierarchy) and runtime squads (Datadog service team tags). Produces SQUAD_MAP.md with confidence and conflict flags. Use when the user asks who owns a repo or service, squad mapping, team ownership, GitLab group to Datadog team reconciliation, or org structure for a multi-repo workspace. Keywords: squad map, ownership, CODEOWNERS, GitLab group, Datadog team, who owns. Not for a full bounded-context/domain map (domain-comprehension) or MR review (pr-review).

0
installs
▌
Test Writer · luckyrjain bundle
Thin orchestration router for test-writing requests that do not resolve to one specialist up front. Classifies the request into one or more complementary test levels, builds an ordered test_plan, then dispatches unit-test-creator, integration-test-creator, contract-test-creator, e2e-test-creator, and/or api-test-creator as required. Preserves each specialist's gates and report verbatim. A single explicitly named level still routes directly to that specialist. Keywords: write tests, generate tests, add test coverage, backfill tests, test this MR/PR/diff.

0
installs
▌
Incident Rca · luckyrjain bundle
RCA and post-incident investigation for outages, error spikes, deploy regressions, and latency incidents in a time window. Keywords: RCA, root cause analysis, postmortem, INC-, P1/P2, SLO breach, on-call, consumer lag, deploy regression. Datadog or KubeSense. Not for MR review, K8s rightsizing, or live remediation. Full phrases: examples.md.

0
installs
▌
Issue Triage · luckyrjain bundle
Classify one or more raw, unscoped incoming issues, bugs, or feature requests: category, severity, duplicate-of (when evidence supports it), and a recommended owning skill or squad. Use when raw issues need triage before anyone acts on them. Keywords: triage these issues, classify these bugs, what category is this, is this a duplicate of another issue, which team owns this issue. Not for a live paging-webhook incident (incident-triage-agent), an already-scoped tracker query to work through (backlog-runner), or a bare ownership lookup with no raw issue to classify (squad-map).

0
installs
▌
Bug Diagnosis · luckyrjain bundle
Diagnose a non-incident bug, test failure, or performance regression: confirm a minimal repro, form candidate root causes and actively try to falsify each one with evidence, report the confirmed root cause. Use when a bug or regression needs its root cause found before it can be fixed, outside a live production incident. Keywords: diagnose this bug, why is this failing, root cause of this test failure, performance regression diagnosis, reproduce this bug. Not for a live production incident with a time window (incident-rca), or applying the fix once the cause is known (loop-task-implementer).

0
installs
▌
Module Design · luckyrjain bundle
Use when one concrete module in an existing repository needs an evidence-backed design for its contract, ownership, seams, dependencies, state, errors, and tests before implementation. Keywords: module design, module boundary, seam design, adapter design, dependency direction, interface design. Not for a multi-module/system design (system-design), or an architecture-wide decision/risk review (architecture-review).

0
installs
▌
Pr Gatekeeper · luckyrjain bundle
Auto-runs pr-review on every push to an open GitLab MR (webhook-triggered), posting inline exactly as pr-review already supports — when pr-review's own posting-confirmation rules allow it. Delegates all review logic to pr-review; only decides whether a given push may auto-post or must route to a human notification instead. Not for interactive, human-typed review requests — those route to pr-review directly. Keywords: push webhook, auto-review, gatekeeper, CI review bot, unattended pr-review.

0
installs
▌
Prd Architect · luckyrjain bundle
Use when rough product ideas, feature proposals, workflows, existing PRDs, or build/no-build questions need a validated PRD, gap review, or readiness assessment. Keywords: PRD, product requirements, should we build, challenge this idea, review PRD, MVP scope, build readiness, feature spec. Not for implementing code (loop-task-implementer), MR review (pr-review), domain mapping (domain-comprehension), or writing tests (test-writer).

0
installs
▌
System Design · luckyrjain bundle
Use when an approved architecture decision or PRD needs to become an implementation-oriented design: components, APIs, events, data model, state machines, consistency, retries, idempotency, capacity, failure strategy, observability, rollout. Keywords: system design, technical design doc, component design, data model, state machine, rollout plan. Not for deciding whether the architecture itself is sound (architecture-review), reviewing an existing API's contract (api-design-review), or writing the PRD (prd-architect).

0
installs
▌
Backlog Runner · luckyrjain bundle
Scheduled queue-management wrapper around loop-task-implementer — pulls N tickets from a Jira/GitHub Issues query, works through them in dependency order overnight, opens a PR per task, never auto-merges. No new Builder/Reviewer/review logic — delegates entirely to loop-task-implementer per task. Not for interactive, human-typed single-task requests — those route to loop-task-implementer directly. Keywords: overnight run, ticket queue, backlog sweep, scheduled multi-task loop, nightly PRs.

0
installs
▌
New Hire Guide · luckyrjain bundle
Personalized onboarding tour for a NAMED new engineer joining a squad — resolves their squad's repos/services via squad-map, runs domain-comprehension (unscoped), and curates the result into a welcome doc. Keywords: new hire, onboarding, new engineer, joining the team, orientation tour. Not for subsystem/domain onboarding with no person named (domain-comprehension) or plain "who owns X" ownership lookups (squad-map).

0
installs
▌
Research Brief · luckyrjain bundle
Investigate a research question against primary sources — repository evidence and, when available, external documentation — and report findings with a cited source and evidence status per claim. Use when a question needs an evidenced, sourced answer rather than a recollection. Keywords: research this, find out whether, what does the documentation say, investigate this question, cited findings. Not for reconstructing this codebase's own current behavior (domain-comprehension), or deciding between options once the facts are known (engineering-decision-discovery).

0
installs
▌
Who Owns X Bot · luckyrjain bundle
Single-shot "who owns X" answer formatted for a Slack slash command or other automated, no-follow-up caller. Delegates the actual ownership lookup entirely to squad-map; this skill only packages the result as one Slack message (resolved / ambiguous / unknown). Not for interactive, conversational ownership questions — those route to squad-map directly. Keywords: /who-owns, Slack slash command, single-shot ownership lookup, bot-facing squad answer.

0
installs
▌
Database Review · luckyrjain bundle
Use when a database schema, migration, or query needs review for indexing, locking, transactions, migration safety, query plans, replication, and partitioning. Keywords: database review, schema review, migration review, index review, query plan, locking, partitioning. Not for the MySQL-to-Postgres rewrite itself (mysql-to-postgres-sql), a general MR review (pr-review), capacity forecasting (capacity-planner), or the blast radius/rollback plan of shipping the change that runs this migration (deployment-risk-review).

0
installs
▌
Domain Modeling · luckyrjain bundle
Build and sharpen a project's domain model as an ongoing discipline during active design or build work: challenge terminology against CONTEXT.md, invent edge-case scenarios, cross-check stated behavior against code, and propose an ADR draft the moment a decision crystallizes. Use when the user is pinning down domain terminology or a ubiquitous language, or when another skill needs live domain vocabulary mid-session. Keywords: domain model, ubiquitous language, glossary, CONTEXT.md, ADR draft, ambiguous terminology, edge-case scenario, ADR record. Not for one-shot reconstruction of an unfamiliar domain from scratch (domain-comprehension), or one concrete module's contract/seam (module-design).

0
installs
▌
Security Review · luckyrjain bundle
Use for a dedicated security review: authentication, authorization, secrets handling, injection, SSRF, tenant isolation, data leakage, cryptography, and dependency exposure. Keywords: security review, authN, authZ, injection, SSRF, tenant isolation, secrets, cryptography review. Not for a general code-quality MR review (pr-review, which escalates here for security-sensitive findings), or a dependency-upgrade CVE sweep (dependency-upgrade-review).

0
installs
▌
API Test Creator · luckyrjain bundle
Generates black-box API test suites (Postman collections, run via Newman) against a real, reachable running API instance. Detects the repo's existing Postman/Newman tooling and collection file(s), writes request/assertion pairs (status code, response schema/fields, headers) for changed endpoints (diff mode) or an explicit endpoint list (backfill mode), chains requests via Postman variables/environment when a flow requires it, runs them, and iterates until green. Every request/response shape traces to real observed usage — never a fabricated payload. Keywords: API tests, Postman, Newman, request/response assertions, black-box API testing, REST endpoint test. Not for an in-process/mocked unit test (unit-test-creator), a real-dependency-seam test via testcontainers (integration-test-creator), a consumer-driven Pact contract (contract-test-creator), or a browser UI journey (e2e-test-creator).

0
installs
▌
Capacity Planner · luckyrjain bundle
Use when historical demand data needs to become forward-looking capacity requirements: RPS, concurrency, CPU, memory, DB, queue, storage, and replica counts. Keywords: capacity planning, capacity forecast, scaling requirements, replica count, headroom. Not for reviewing current resource rightsizing against live metrics (k8s-overprovisioning-datadog) or a performance code review (performance-review).

0
installs
▌
E2e Test Creator · luckyrjain bundle
Generates full user-journey end-to-end tests through a real browser UI using Playwright, Cypress, or Selenium/WebDriver. Detects the repo's browser test tooling and layout convention, infers journeys from changed routes/pages (diff mode) or takes explicit journey descriptions (backfill mode), writes tests asserting only on user-visible outcomes (never internal DOM/state), runs them against a reachable app instance, and iterates until green. Keywords: e2e tests, end-to-end, browser test, Playwright, Cypress, Selenium, user journey, click-through test. Not for below-the-UI service-seam tests (integration-test-creator), isolated mocked unit tests (unit-test-creator), or consumer/provider contract agreements (contract-test-creator). Web browser flows only — not API/CLI black-box journeys.

0
installs
▌
API Design Review · luckyrjain bundle
Use when a REST, GraphQL, gRPC, or async-event API design or contract needs review for compatibility, pagination, idempotency, error semantics, versioning, authorization, and rate limiting. Keywords: API design review, API contract review, breaking change, API versioning, pagination design. Not for a full merge-request code review (pr-review), database schema review (database-review), or implementation-level system design (system-design).

0
installs
▌
Local Diff Review · luckyrjain bundle
Review the changes since an arbitrary fixed point (commit, branch, tag, or merge-base) — not a live PR/MR — along two independent axes: Standards (this repo's documented conventions) and Spec (does the diff match a supplied issue/ticket text). Use when the user wants a local or uncommitted diff reviewed before it becomes a PR/MR. Keywords: local diff review, review since commit, review against branch, uncommitted diff review, review this branch. Not for a live PR/MR by number (pr-review), or existing-codebase architecture friction (codebase-architecture-review).

0
installs
▌
Resilience Review · luckyrjain bundle
Review a proposed design or current implementation for timeout budgets, retries, circuit breaking, load shedding, backpressure, queues, idempotency, downstream failures, partial failures, and recovery. Keywords: resilience review or failure-mode review outside a live incident. Not for incident diagnosis, capacity forecasting, Kubernetes rightsizing, or generic PR review.

0
installs
▌
Unit Test Creator · luckyrjain bundle
Generates isolated, fast unit tests — function/class-level, every external dependency mocked or stubbed — for a target repository. Detects the repo's own test framework (pytest, Jest/Vitest/Mocha, Go testing, JUnit via Maven/Gradle, RSpec/Minitest, xUnit/NUnit/MSTest, cargo test), writes tests for changed code (diff mode) or an existing coverage gap (backfill mode), runs them, and iterates until green. Keywords: unit tests, mock externals, isolated test, fast test, TDD helper. Not for tests that need a real adjacent dependency (integration-test-creator), consumer/provider contract agreements (contract-test-creator), or full browser user journeys (e2e-test-creator).

0
installs
▌
Performance Review · luckyrjain bundle
Use when code, a query, or a service needs review for algorithmic complexity, database behavior, N+1 queries, caching, memory, concurrency, connection pools, and downstream fanout. Keywords: performance review, N+1, slow query, cache design, concurrency review, connection pool sizing. Not for turning demand into capacity numbers (capacity-planner) or reviewing schema/index design directly (database-review).

0
installs
▌
Tech Debt Assessor · luckyrjain bundle
Use when a backlog of tech debt items needs ranking by business impact, engineering drag, operational risk, and effort. Keywords: tech debt, debt prioritization, debt backlog, engineering drag, refactor prioritization. Not for planning a specific migration program (migration-program-manager, which this skill can escalate to) or a cost/rightsizing sweep (cost-optimization-sprint-planner).

0
installs
▌
Architecture Review · luckyrjain bundle
Use when a PRD, proposed design, or architecture diagram needs a validated architecture decision before implementation begins — risks, scale limits, failure modes, security posture, operability, and alternatives considered. Keywords: architecture review, ADR, architecture decision record, design review, should we build it this way, scale limits, failure modes. Not for authoring the PRD itself (prd-architect), implementation-level component/API/data-model design (system-design), or reviewing an already-merged PR's code (pr-review).

0
installs
▌
Weekly Squad Digest · luckyrjain bundle
Scheduled digest combining migration-program-manager's and cost-optimization-sprint-planner's own rollup JSON outputs into one squad-grouped report — never re-runs either aggregator, never re-derives squad/status/priority. Keywords: weekly squad digest, combined rollup, org-wide status roundup. Not for a fresh migration or cost rollup (migration-program-manager / cost-optimization-sprint-planner directly) or a plain ownership lookup (squad-map).

0
installs
▌
Domain Comprehension · luckyrjain bundle
Build a verifiable, evidence-backed representation of a business domain and an as-built PRD for the in-scope service(s) and/or domain. Executable source code is primary truth; runtime telemetry validates behavior, not intent. Keywords: subsystem onboarding, multi-repo architecture ground truth, bounded-context mapping, current-state requirements reconstruction, and engineering-leader summaries. Not for ownership lookup only (squad-map), MR review (pr-review), or future-state product specification from an idea (prd-architect).

0
installs

46 published skills