Habilidades research
📦

research

Revisão do conteúdo r1 Seguro ⚙️ Comandos externos

Delegate Broad Codebase Research

Large investigations can flood the primary agent context with irrelevant files and logs. This skill delegates read-only research and returns concise conclusions with cited evidence.

Suporta: Claude Codex Code(CC)
🥉 77 Bronze

Instalar com meu Agente

Copie esta solicitação para o seu Agente. Ela inclui a página canônica do Skill e o manifesto.

Solicitação do agente
Review the Skillstore skill "research" from https://skillstore.io/skills/warpdotdev-research.md and its manifest at https://skillstore.io/api/skills/warpdotdev-research/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

Seu Agente ainda deve apresentar o plano e solicitar qualquer confirmação exigida pela política de segurança.

Recursos legíveis por agentes

Use estes links quando um agente de IA, crawler ou script precisar de contexto limpo em vez de ler a página completa.

Detalhe em Markdown GET /skills/warpdotdev-research.md manifest assinado GET /api/skills/warpdotdev-research/manifest lockfile assinado GET /api/skills/warpdotdev-research/lockfile

Testar

Usando "research". Find the root cause of this failing integration test and cite the relevant implementation and test locations.

Resultado esperado:

  • Root cause: The retry handler converts the timeout into a permanent failure.
  • Evidence: The report identifies the handler and failing test locations.
  • Caveat: External service behavior was not reproduced locally.

Usando "research". Summarize this large pull request, including behavior changes, risks, and test coverage.

Resultado esperado:

  • Change: The request pipeline now validates credentials before loading account data.
  • Risk: Existing clients with incomplete credentials may fail earlier.
  • Coverage: Unit tests cover validation, but migration behavior remains untested.

Usando "research". Explain how session recovery works across the repository and identify the main call sites.

Resultado esperado:

  • Flow: Recovery starts after connection loss, restores persisted state, and reconnects active sessions.
  • Evidence: The summary lists the coordinator, persistence layer, and reconnect entry points.
  • Unknown: Concurrent recovery behavior requires runtime verification.

Auditoria de segurança

Seguro
v1 • 04/09/2026 Abrir o relatório versionado

All five static findings are false positives. Four are Markdown inline-code spans, while line 36 discusses subagent costs without performing reconnaissance. No prompt injection, command execution, or semantic security issue was found.

1
Arquivos analisados
72
Linhas analisadas
0
Itens de revisão
0
Falsos positivos ignorados

Fatores de risco

⚙️ Comandos externos (4)
SKILL.md:27 SKILL.md:33 SKILL.md:44 SKILL.md:64
A auditoria estática e semântica concluída mais recente não detectou achados de segurança confirmados. Isso não comprova que a skill não tenha efeitos colaterais.
Auditado por: codex
Compartilhar e citar este relatório

Compartilhe o relatório de avaliação versionado, o badge neutro, o cartão incorporável e as citações. A Skillstore apresenta evidências sem decidir se este Skill é seguro.

Abrir o relatório versionado
Avaliação de Segurança

Copiar link do relatório

https://skillstore.io/skills/warpdotdev-research/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Badge Markdown

[![Skillstore security assessment](https://skillstore.io/badges/skills/warpdotdev-research/security.svg)](https://skillstore.io/skills/warpdotdev-research?utm_source=security_passport_badge)

Badge HTML

<a href="https://skillstore.io/skills/warpdotdev-research?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/warpdotdev-research/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Cartão incorporável

<iframe src="https://skillstore.io/embed/skills/warpdotdev-research.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>
Citações acadêmicas (APA · BibTeX · CFF)

Citação APA

warpdotdev. (2026). research security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/warpdotdev-research/audits/1

Citação BibTeX

@techreport{warpdotdev-warpdotdev-research-2026, author = {warpdotdev}, title = {research security audit report (audit version 1)}, institution = {Skillstore}, year = {2026}, number = {1}, url = {https://skillstore.io/skills/warpdotdev-research/audits/1}, note = {Author version unspecified} }

CITATION.cff

cff-version: 1.2.0 message: "If you use this Skill, cite its author and this versioned security audit report." title: "research security audit report (audit version 1)" version: "unspecified" type: report authors: - name: "warpdotdev" date-released: "2026-09-04" url: "https://skillstore.io/skills/warpdotdev-research/audits/1" identifiers: - type: other value: "skillstore:warpdotdev-research:audit:1" description: "Skillstore immutable audit report identifier"

Comparar variantes

2 variantes instaláveis

Cada autor permanece como um skill instalável separado. A variante recomendada é classificada pelas evidências do Skillstore.

Por que esta variante está em primeiro

Maior uso no Skillstore
mattpocock Recomendado

mattpocock-research

Pontuação do Skillstore 77
Confiança das evidências Médio
Uso no Skillstore 11
Atualizado

2026-09-09

warpdotdev Atual

warpdotdev-research

Pontuação do Skillstore 77
Confiança das evidências Médio
Uso no Skillstore 5
Atualizado

2026-09-09

Pontuação do Skillstore

Por que esta pontuação Confiança das evidências: Médio
55
Arquitetura
85
Manutenibilidade
87
Conteúdo
65
Comunidade
83
Conformidade com a especificação

O que você pode criar

Trace a complex test failure

Delegate inspection of long CI logs and related code, then receive the likely cause with supporting locations.

Summarize a large pull request

Review a substantial diff through a subagent and receive the main changes, rationale, risks, and evidence.

Survey an unfamiliar architecture

Map how a component or API is used across a repository without loading every inspected file into primary context.

Teste estes prompts

Answer a repository question 
Investigate {question} in {repository_path}. Work read-only. Return a concise answer with exact file paths, symbols, and important caveats.
Diagnose a failure 
Analyze {log_or_test_failure} and trace the relevant code. Identify the root cause, cite supporting locations, and note anything not verified.
Review a large change 
Research {pull_request_or_diff}. Summarize behavioral changes, affected components, design rationale, risks, and tests. Support each major conclusion with file locations.
Coordinate independent investigations 
Investigate these independent areas in parallel: {area_one}, {area_two}, and {area_three}. Synthesize the results, preserve evidence, and identify cross-area uncertainties.

Boas práticas

  • Provide an exact question, repository path, and suspected files or symbols.
  • Keep investigations read-only unless the task explicitly requires changes.
  • Request concise conclusions, exact evidence locations, caveats, and unresolved questions.

Evitar

  • Do not delegate a single lookup or a few known file reads.
  • Do not delegate when the primary agent must edit the same files immediately.
  • Do not split one coherent investigation into artificial parallel tasks.

Perguntas frequentes

When should I use this skill?
Use it when investigation requires many files, long logs, large diffs, or broad repository surveys.
When should I inspect files directly?
Inspect files directly for small lookups or when you need their contents for immediate edits.
Does the skill modify repository files?
The workflow is read-only by default. A task must explicitly request changes before a subagent should modify files.
Can it run several investigations at once?
Yes. Parallel subagents are appropriate only when the questions are independent and do not require shared intermediate findings.
What should a research report include?
It should include a direct answer, exact evidence locations, relevant symbols, caveats, and unresolved questions.
Which AI tools support this skill?
The report declares support for Claude, Codex, and Claude Code. Local subagent availability still depends on the host environment.

Detalhes do desenvolvedor

Licença

MIT

Revisão da Skillstore

r1

Aviso de versão

O autor não declarou uma versão.

Repositório

https://github.com/warpdotdev/common-skills/tree/b811c24365ae505bfc9646458957b886e29110b5/.agents/skills/research

Referência

e5464e662405de6361fdde6b984f9ea865635f64

Atualidade da manutenção

08/09/2026

Uso

1 downloads · 0 visualizações

Estrutura de arquivos

📄 SKILL.md