Explicit Research Mode (re)

Trust Score 85/100

Temporarily bias the agent toward a research-heavy workflow: plan sub-questions, retrieve sources, synthesize with explicit citations, and verify before impleme

triggers:/reresearch-heavyextra-researchsource-checkdeep-research

GitHub SKILL.md

What it does

This skill adds an explicit extra-research mode to the agent. When enabled, the agent prioritizes source retrieval, source-backed justification, and a three-pass research pattern (plan sub-questions, retrieve sources and follow leads, then synthesize). It emphasizes local code inspection first, uses web sources for current facts, and demands explicit citation of supporting evidence or stated conflicts. The mode is intentionally lightweight — it nudges the agent toward diligence without replacing normal orchestration.

When to use it

Use this skill when the user explicitly requests a research-focused pass (e.g., the user types /re) or when a task clearly requires careful source-checking before coding or deciding. Ideal for version-sensitive questions, ambiguous library behaviors, controversial claims, or when external validation will materially change the outcome. Avoid for trivial or quick answers.

What's included

Scripts: none declared in the repo location for this skill
References: none declared in the skill directory
Instructions: a concise 3-pass research routine (plan sub-questions, retrieve sources with 1-2 second-order leads, synthesize after resolving conflicts), local-first checks, source citation rules, guidance on avoiding forced phases, and a short finish checklist (decision, source-backed justification, changed code/config, tests run, remaining unknowns).

Compatible agents

Likely compatible with Codex-style orchestrations and CLI-driven agents (OpenAI Codex, Copilot/Copilot-style tooling, and similar agents that support web retrieval).

Audit Summary

Pure instructional skill with no scripts — a research-mode override for Codex agents. Clean SKILL.md with clear defaults, avoid-sections, and finish guidelines. No executable code to test; static analysis only. Well-structured but niche in applicability.

Watch Out

Only useful for Codex/agents with web search and Context7 access
No scripts to validate — purely declarative

Notes

No security concerns. Simple, well-written instructional skill. Lacks scripts or references/ but doesn't need them for its purpose.

Information

Repository: dgk-gpt
Stars: 50

Trust Score

Overall85

Security100

Code Quality72

Architecture65

Usefulness55

More from dgk-gpt

Verify, Commit and Push (CP)

A secure local gate that verifies changes, stages specific files, and pushes to the current branch for Codex agents.

Related Skills

Retriever (Membrane CLI)

Integrate Retriever search and discovery via the Membrane CLI — manage connections and run queries across Retriever data sources.

Monad Development

Practical developer skill for deploying and verifying smart contracts, configuring Foundry projects, and integrating frontends (viem/wagmi) on the Monad blockch

Maestro Orchestrate

Orchestrates complex engineering workflows with enforced design and approval gates, child-agent delegation, and session state management.

Run Execute

Orchestrates execution of work items across modes (autopilot, confirm, validate) with scripted init/complete tooling, plan/test/report artifacts and strict gati

arxiv2md

Convert arXiv papers into clean, LLM-ready Markdown via a REST API or CLI for fast reading, summarization, and downstream processing.

Ip2Location IO Automation (Composio)

Automate Ip2location IO tasks via Composio's Rube MCP toolkit — discover tools, verify connections, and execute schema-compliant workflows safely.

Codex Collab

Bridge Claude and Codex to run tasks, perform code reviews, and run background jobs for research or implementation workflows.

Scite Database (Smart Citations)

Query Scite.ai Smart Citations by DOI to report supporting, contrasting, and mentioning citation counts and help assess a paper's reliability.

Back to Skills