core-skills

Plugin core-skills from marketplace sharekit-profile in LucasSantana-Dev/sharekit-profile. Commands read from the repo's manifest — 3 days ago.

Skills (51)

• adt-rag-drift

  Detect and fix stale chunks (files that changed or were deleted since last indexing)

• app-teardown

  Use when asked \"faz um teardown do app X\", \"analisa esse app e mapeia ideias\", \"what can we steal from X\", \"reverse-engineer this product\", \"competitive teardown\", \"o que dá pra aproveitar do app Y\". Reverse-engineers a live application (web app, SaaS, mobile) by walking its real flows i

• audit-deep

  Composite project health check (testing, config, hooks, security, MCP, plugins) into one severity-ranked report. Use weekly or before major releases.

• backlog

  Composite backlog builder for one repo: scores findings by ROI, groups themes, creates GitHub issues on a Project board. Use to build a backlog or plan sprints.

• caveman

  Ultra-compressed reply mode: drops filler, articles and pleasantries, keeps full technical accuracy (about 75% fewer tokens). Use on \"caveman mode\", \"less tokens\", \"be brief\" or /caveman.

• changelog-update

  Update CHANGELOG.md (Keep a Changelog): promote [Unreleased] to a versioned section. --bump also infers semver, syncs monorepo versions, refreshes lockfiles and opens the release-prep PR. For repos without release-please.

• client-offboard

  Keep the technical and behavioral lessons from a client engagement and remove that client's business knowledge from memory and the RAG index. Modes - harvest (batch-promote pending lessons to general memory), promote-now NOTE (one lesson), offboard SLUG (final harvest, purge, verification). Use for

• code-teardown

  Use when asked \"faz um teardown desse repo\", \"analisa esse repositório pra ideias\", \"o que dá pra aproveitar desse projeto open-source\", \"evaluate X repo for insights\", \"what can we learn from this codebase\". Reverse-engineers an external repository or codebase (competitor OSS, reference i

• codebase-memory

  Use the codebase knowledge graph for structural code queries. Triggers on: explore the codebase, understand the architecture, what functions exist, show me the structure, who calls this function, what does X call, trace the call chain, find callers of, show dependencies, impact analysis, dead code,

• context-pack

  Build a compact, task-aware context bundle (code, standards, plans, RAG hits) before large changes, reviews or unfamiliar repos. Skip when the task touches one known file or fits one grep.

• debate

  Run a structured multi-agent debate on a decision: independent positions, rebuttal, reconciled synthesis. Use to weigh alternatives before deciding.

• dep-sweep

  Batch-process Dependabot/Renovate PRs by risk: auto-merge safe ones, surface risky ones for review. Use when bot PRs pile up.

• dispatch

  Parallelize independent audit or evidence-gathering work into 2+ concurrent agent tracks and reconcile into one verdict; also dispatches to <homelab> over SSH.

• efficiency-advisor

  Analyze a proposed multi-agent workflow, Workflow() script, or active session for token waste and time bottlenecks. Catches model-tier mismatches, sequential work that should be parallel, re-read waste, and context bloat — scores each fix by impact with estimated savings. Use before spawning large a

• frontend-reference-hunt

  Search Dribbble, Mobbin, Awwwards for UI references into a board. Use for \"busca referências\", \"find design references\", \"moodboard\". Does not build UI.

• gh-fix-ci

  Diagnose red, flaky or blocked PR checks: isolate the first real blocker from noise, read Actions logs, and fix failing checks on your own PR branch. Handles Buildkite and SonarCloud by reporting. Use when CI is red before merge.

• graphify

  Answer codebase, architecture and file-relationship questions via a graphify knowledge graph (query/path/explain), especially when graphify-out/ exists.

• grill-with-options

  Interview skill with four modes: options (default, bounded 2-4 option AskUserQuestion forks), open (Q plus GUESS), docs (vs CONTEXT.md and ADRs), explore (2-3 approaches, no code). Use for grill me, help me decide, brainstorm.

• handoff

  Compact the current conversation into a handoff document for another agent to pick up.

• humanizer-pt-br

  Remove traços de IA de textos em português brasileiro, deixando-os naturais. Gatilhos: humanizar texto, remover IA, texto robótico, parece IA, soar natural.

• humanizer

  Rewrite AI-sounding text so it reads like the writer without changing what it says. Use when editing or reviewing prose for AI tells: not-X-but-Y contrasts, one-line closers, staged openers, forced triads, dashes everywhere, inflated claims, sales language, stock AI words, bold labels, or filler. Ba

• knowledge-loop

  Persist knowledge: recall, sync-memories, rag-curate, handoff. Modes: checkpoint, close (wrap up, sign off, save and stop, close session, --ship). Also \"what did we decide\", \"remember this\".

• loop-engineer

  Designs and builds autonomous agent loops. Use for "design a loop", "build an agent loop", "I want this to run automatically", or self-running feedback cycles.

• loop

  Detect when the agent repeats the same action without progress and break the cycle. Use when stuck, "you're going in circles", or an error recurs 3+ times.

• memory-prune

  Audit project memory for stale entries (merged PRs, fixed bugs, missing files); propose archival or deletion. Read-only. Use monthly or on memory/code conflict.

• next-priority

  Decide the highest-value safe thing to do right now in the active repo or workspace.

• orchestrate

  Coordinate multi-agent teams and multi-step or multi-repo work across plans, worktrees, and parallel investigations. Use for large parallelizable tasks.

• parallel-work-coordinator

  Governance layer for small (≤5-unit) parallel batches that MIX read-only and write-capable agents — autonomy-tier gating (T0-T3) and worktree/merge discipline for the writes. Thin wrapper over dispatch; use parallel-investigate instead when every unit is read-only. Invoke on "do all these", "audit a

• plan

  Create a structured implementation plan with phases, file paths, and checks. Use for "plan this", "create a plan", "break this down", "phase this work".

• pr-merge-readiness

  Aggregate every PR merge signal (CI, reviews, conflicts, scans, bots) into one verdict, MERGE / WAIT / FIX. Use before merging. --batch [N...] gives a read-only verdict table for several PRs or all your open PRs.

• prototype

  Build a throwaway prototype before committing to a design: terminal app or UI variations. Use for \"prototype this\", \"let me play with it\", \"try a few designs\".

• rag-curate

  Improve RAG corpus quality by adding missing docs, rewriting weak chunks, filling gaps. Use after a diagnostic flags a gap or when the index returns poor hits.

• recall

  One-shot semantic lookup in the local RAG index. Use for "what did we decide about X", "where did we hit this bug before", "is there a memory note for Y".

• refactor

  Surgical refactoring without behavior change. Use for "refactor this", "clean up this code", "this is messy", "extract functions", "fix code smells".

• research-and-decide

  Decide with three modes: quick (small reversible pick), full (research, critic, plan, ADR), now (break analysis paralysis: 'just decide'). Use to evaluate a library, pattern or architecture, or force a call.

• resolve-before-asking

  Answer a would-be clarifying question yourself when researchable, reversible, not T3. Never for destructive, prod, money or other-author-PR actions: ask.

• resume

  Rehydrate the current task from handoffs, plans and git. Triggers: resume, continue, what was I doing, where was I, wake up. Default is full; --brief gives a 4-section brief (~800 tokens).

• retomar

  Alias for the resume skill that the Claude desktop app does not intercept (its native /resume opens the session picker). Use for /retomar, /retomar N, retomar handoff.

• review

  Review changes since a fixed point (commit, branch, tag) on two axes, Standards and Spec, in parallel. Use to review a branch, PR or WIP, or "review since X".

• secure

  Security-first pass over config, credentials, dependency risk and unsafe code patterns. Covers OWASP Top 10, infra config (Terraform/K8s/Docker), CORS/CSRF.

• session-bootstrap

  Start-of-day routine: resume --brief, next-priority, PR queue, context-pack. Use for "what was I doing", "what''s next", "load context".

• ship-it

  Take a merged PR to production: changelog-update --bump, tag/release, deploy, post-deploy verify. --from tag starts at the tag phase. For "live in prod, verified". Unmerged work: use merge-confidently.

• spec-driven-develop

  Default entry for non-trivial build/add/fix/refactor: specify, plan, tasks, implement, verify. Not if a narrower composite (hotfix etc) fits; skip <3 files.

• streamer-mode

  Behavioral safety mode for when the user is live streaming or screen sharing — never print secrets, env vars, tokens, IPs, hostnames, or terminal-sensitive data on screen; substitute safe command alternatives; defer security findings to a private file instead of announcing them on stream. Use when t

• sync-memories

  Sync durable project or session knowledge into the available memory systems

• sync-sharekit-profile

  Sync ~/.claude/ (skills, hooks, standards, CLAUDE.md, agents) into the public sharekit profile repo, sanitize personal paths/identity, scan for secrets, show diff, and commit+push. Invoke whenever: "sync my sharekit profile", "update sharekit", "push my latest skills to sharekit", "sharekit is out o

• tdd

  Short alias for test-driven development. Use for red-green-refactor, tracer bullets, behavior-driven testing, or TDD workflows.

• test-driven-development

  Use when implementing any feature or bugfix, before writing implementation

• three-man-team

  Parallel Architect (Opus), Builder, Reviewer (Sonnet) dispatch for complex multi-phase tasks. Use for 3+ phase features, CRITICAL work, or PRs over 5 files.

• verify-before-done

  Pre-ship gate: full matrix (lint, build, tests, coverage, sonar, CI, sentry) or quick mode (narrowest checks, PASS/FAIL/PARTIAL). Use for "is this ready?", verify, run checks, validate this.

• xp

  Extreme Programming for AI-human pair development. Use when: pair programming with your AI, building incrementally with tests, working with clear role division, or the user mentions XP/YAGNI/simple design/continuous refactoring.

Similar plugins