By jryanvieira
Framework de review de Pull Requests com agentes de IA. 6 agentes especialistas, grounding obrigatório, anti-hallucination, e suporte a integrações externas.
All checks performed by `/trust doctor`, with pass/fail criteria and output format.
Load this file only when `--fix` flag is passed to `/trust doctor`.
In enterprise profile, TRUST uses a **hierarchical layer structure**:
```
In solo profile, TRUST uses your **second brain as primary grounding source**.
Reviews pull request diffs for API contract violations defined in the project's contract spec and grounding documents. Use when analyzing code changes for breaking changes, schema regressions, missing versioning, backward-incompatible field additions, or deviations from documented API contracts. Always operates within project-specific rules. Activated by the TRUST orchestrator during PR review execution.
Reviews pull request diffs for code convention violations defined in the project's conventions grounding documents. Use when analyzing code changes for naming violations, logging pattern deviations, error handling anti-patterns, file structure violations, or any code style rule documented in the project's conventions guide. Always operates within project-specific rules. Activated by the TRUST orchestrator during PR review execution.
Reviews pull request diffs for data model and database migration violations defined in the project's data policy and grounding documents. Use when analyzing schema migrations, model changes, index additions/removals, constraint changes, or data integrity rules. Always operates within project-specific rules. Activated by the TRUST orchestrator during PR review execution.
Performs the Meta-Review (anti-hallucination) phase of the TRUST framework. Validates every finding from all specialist agents against the actual diff content. Verifies that evidence_quote is literal, rule_source resolves to real grounding, and claim does not exceed what the evidence supports. Use when the orchestrator reaches Phase 5. Intercepts hallucinated findings before they reach the developer.
Reviews pull request diffs for performance violations defined in the project's performance policy and grounding documents. Use when analyzing code changes for N+1 queries, missing indexes, unbounded queries, missing timeouts, synchronous blocking operations, or memory hotspots. Always operates within project-specific rules. Activated by the TRUST orchestrator during PR review execution.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Trustable Reviews via Universal Skills & Tooling O review de IA em que dá pra confiar.
TRUST é um framework de review de Pull Requests com agentes de IA, baseado em sete pilares que separam um review confiável de um que devs ignoram. Ele roda local no Claude Code, é agnóstico de stack, e vive desacoplado do repo do produto.
TRUST é para times que já tentaram AI code review e desistiram. Os sintomas conhecidos:
"A IA aponta coisas que contradizem nossos ADRs." "70% dos alertas são ruído — a gente ignorou tudo." "Ela inventa imports e cita linhas que não existem." "Não temos como saber se o review checou tudo."
Se você reconhece pelo menos dois desses, TRUST foi desenhado para você.
# 1. Instala o framework (uma vez por máquina)
/plugin marketplace add jryanvieira/trust
/plugin install trust@trust
# 2. Clona o setup do seu time (uma vez por dev)
git clone github.com/empresa/payments-team-trust ~/work/payments-team-trust
export TRUST_SETUP_PATH=~/work/payments-team-trust
# 3. Trabalha no repo do produto normalmente, sem alterações
cd ~/work/api-payments
git checkout -b feat/PAY-123
# 4. Roda o review
/trust review-pr feat/PAY-123
O framework auto-descobre o setup pela env var, casa o repo atual com um target.yaml, carrega o grounding, executa as 8 fases com DoD estrito, e gera um REVIEW.md fora do repo do produto.
Detalhes em docs/03-onboarding.md.
┌─────────────────────┐ ┌──────────────────────┐ ┌─────────────────────┐
│ 1. Framework TRUST │ │ 2. Setup do Time ★ │ │ 3. Repo do Produto │
│ (este repo) │ ─→ │ payments-team-trust │ ─→ │ api-payments │
│ o motor │ │ config + regras │ │ 100% LIMPO │
└─────────────────────┘ └──────────────────────┘ └─────────────────────┘
instalado uma vez clonado uma vez nunca tocado pelo TRUST
Vinculação por uma única env var (TRUST_SETUP_PATH). Detalhes em docs/02-arquitetura.md.
A fonte da verdade do projeto está organizada por tópico para você navegar pelo que importa agora:
| Doc | Para que serve |
|---|---|
| docs/01-conceitos.md | Os 7 pilares, profiles operacionais (Pilot/Team/Solo/Enterprise), filosofia |
| docs/02-arquitetura.md | 3 repos, 8 fases, sources plugáveis, hierarquia de camadas |
| docs/03-onboarding.md | Instalação, init, doctor, primeiro review passo a passo |
| docs/04-skills-spec.md | Como cada skill segue o padrão agentskills.io oficial |
| docs/05-glossario.md | Vocabulário do TRUST (DoD, halt, grounding, target, layer, run-id) |
| CHANGELOG.md | Histórico de versões com plano de entrega detalhado por fase |
Estes princípios estão no DNA do framework. Mudanças que violem qualquer um destes devem ser rejeitadas.
rule_id rastreável até uma seção do grounding.✅ v2.3.0 — Plugin Distribution. TRUST agora é instalável diretamente via /plugin marketplace add + /plugin install, sem configuração manual. Todos os commands têm frontmatter description: e o manifesto .claude-plugin/ está completo. Veja o CHANGELOG para o histórico completo de versões.
npx claudepluginhub jryanvieira/trust --plugin trustSpec-driven development framework with parallel execution, pluggable issue trackers, and a learning loop. Covers the full lifecycle: product discovery → spec → design → plan → execute → PR → review.
Agent Runtime Goal Optimizer — otimiza harnesses de coding agents automaticamente
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Comprehensive startup business analysis with market sizing (TAM/SAM/SOM), financial modeling, team planning, and strategic research
Next.js development expertise with skills for App Router, Server Components, Route Handlers, Server Actions, and authentication patterns