research-skills
v0.3.0AI/ML precedent investigation, problem decomposition, result interpretation, cue recombination, hypothesis development, and iterative research, with independent records, pilots, reports, mirrors, and GPU execution.
By hemaher00 GitHub starsUpdated 1 hour ago
Directory evidence
- Runtimes
- Codex
- Parsed components
- 11 skill or MCP entries
- Source updated
- Oct 1, 2026
- Manifest status
- Canonical path parsed
The directory validates manifest shape and source location. It does not execute the plugin or provide a security endorsement. Review the indexing methodology →
Install research-skills for Codex
codex plugin marketplace add hemaher0/research-skills
codex plugin marketplace upgrade hemaher0-research-skills
codex plugin add research-skills@hemaher0-research-skillsPaste and run these commands in a terminal with Codex. They add and refresh the hemaher0-research-skills catalog, then install this plugin.
Compatibility: the page URL and API slug “research-skills” remain stable.
- Codex:
research-skills@agent-plugin-marketplace→research-skills@hemaher0-research-skills
The installer fetches third-party code from the source repository shown on this page. This directory validates manifest structure and source location, but does not perform a security audit; review the manifest, components, and source before installing.
Get the source manually
git clone https://github.com/hemaher0/research-skillsClone the source repository, then follow its setup instructions to add the plugin to a compatible client. The plugin root is plugins/research-skills/.
Plugin files
├── .codex-plugin/plugin.json├── skills/digging/SKILL.md├── skills/evidence-based-research/SKILL.md├── skills/experiments/SKILL.md├── skills/exploration/SKILL.md├── skills/interpretation/SKILL.md├── skills/notion-mirror/SKILL.md├── skills/pilot/SKILL.md├── skills/probing/SKILL.md├── skills/reinterpretation/SKILL.md├── skills/report/SKILL.md└── skills/run-gpu/SKILL.md
Included Skills11
Use when a specific AI/ML research problem, assumption, method, or unexpected phenomenon needs decomposition, implementation inspection, or targeted reproduction to understand it.
Use when carrying an AI/ML inquiry through evidence-led iterations and choosing or revising the next attempt, explanation, method, or question.
Use when creating or continuing a persistent AI/ML experiment record and managing its protocol, lineage, execution state, and evidence. Not for standalone result investigation, a standalone pilot or report, or design discussion without a requested record.
Use when investigating scientific precedents, methods, implementations, or conflicting prior evidence for an AI/ML research question.
Use when analyzing existing AI/ML research results, choosing meaningful transformations or comparisons, and judging what the evidence supports or leaves unresolved.
Use when the user explicitly requests mirroring a completed local AI/ML experiment document to its configured Notion destination. Not for experiment completion, report writing, or result investigation alone.
Use when planning or running a bounded AI/ML research pilot to obtain new evidence under representative conditions. Works without an experiments record. Not for inspecting existing results, routine software tests, or code changes alone.
Use when developing or revising a provisional AI/ML research hypothesis from evidence and deciding which observations could clarify its assumptions, consequences, or alternatives.
Use when multiple AI/ML research cues need recombination or a change of representation to reveal a different connection, concept, or research question.
Use when presenting AI/ML research findings in a requested answer, report, or existing record. Use interpretation for result analysis; reporting does not require a completed study.
Use when a Git repository with a shared uv or Conda environment requires NVIDIA GPU or CUDA diagnostics, torch.cuda tests, CUDA pytest cases, NCCL, nvidia-smi, or torchrun from its main checkout or linked worktrees.
Plugin manifests1
{
"name": "research-skills",
"version": "0.3.0",
"description": "AI/ML precedent investigation, problem decomposition, result interpretation, cue recombination, hypothesis development, and iterative research, with independent records, pilots, reports, mirrors, and GPU execution.",
"author": {
"name": "hemaher0",
"url": "https://github.com/hemaher0"
},
"repository": "https://github.com/hemaher0/research-skills",
"keywords": [
"research",
"exploration",
"hypotheses",
"interpretation",
"experiments",
"pilot",
"report",
"notion",
"machine-learning",
"gpu"
],
"skills": "./skills/",
"interface": {
"displayName": "Research Skills",
"shortDescription": "Evidence-led AI/ML inquiry and research operations",
"longDescription": "Six reasoning skills for investigating precedents, decomposing problems, interpreting results, recombining research cues, developing provisional hypotheses, and advancing an inquiry through evidence-led iterations. Independent skills manage persistent experiment records, representative pilots, research reports, requested Notion mirrors, and repository-scoped shared-uv GPU execution.",
"developerName": "hemaher0",
"category": "Productivity",
"capabilities": [
"Read",
"Write"
],
"defaultPrompt": [
"Investigate research cues and choose an informative next attempt within this inquiry's scope.",
"Develop a provisional hypothesis from the available evidence and expose its assumptions.",
"Analyze existing research results and state what remains unexplained.",
"Manage a requested persistent experiment record or plan representative new-evidence execution.",
"Inspect this repository's shared uv environment and the GPUs allocated on this server."
]
}
}For maintainers
If you maintain this plugin, link to this source-backed listing from your README so users can review its manifest and indexed components.
[research-skills on Agent Plugins Marketplace](https://pluginsmp.com/plugins/research-skills)