Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/HirogaKatageri/hirokataWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/plugins/hirogakatageri/hirokata/research)<a href="https://agentmods.dev/plugins/hirogakatageri/hirokata/research"><img src="https://agentmods.dev/badge/plugins/hirogakatageri/hirokata/research/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/plugins/hirogakatageri/hirokata/research"><img src="https://agentmods.dev/badge/plugins/hirogakatageri/hirokata/research.svg" alt="Reviewed on agentmods" width="80" height="20"></a>Grade A, and why
research scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
{
"name": "research",
"version": "1.0.0",
"description": "PhD-level multi-perspective research inspired by Stanford's STORM. The 'storm research' skill orchestrates five expert persona agents (practitioner, skeptic, economist, historian, academic), maps their contradictions, synthesizes a cited briefing, and red-teams the result. Say 'storm research <topic>' to start.",
"author": {
"name": "Gian Patrick Quintana",
"email": "[email protected]"
},
"homepage": "https://github.com/HirogaKatageri/hirokata/tree/main/plugins/research",
"repository": "https://github.com/HirogaKatageri/hirokata",
"license": "MIT",
"keywords": ["research", "storm", "multi-perspective", "agents", "synthesis", "peer-review", "deep-research"]
}
What it installs
The manifest is a name and a version. 4 skills, 8 agents travel with it, and installing the plugin installs all of them — 1,201 tokens a session between them. Each is measured on its own page, and each can be installed alone.
- Skill storm-research A 141 tokens
- Skill contradiction-map A 120 tokens
- Skill multi-perspective-scan A 141 tokens
- Skill research-peer-review A 144 tokens
- Agent academic A 77 tokens
- Agent contradiction-mapper A 80 tokens
- Agent economist A 78 tokens
- Agent historian A 81 tokens
- Agent peer-reviewer A 94 tokens
- Agent practitioner A 89 tokens
- Agent skeptic A 79 tokens
- Agent synthesizer A 77 tokens
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 14 lines scan A fff90c5ef49a
research is a plugin published in the GitHub repository HirogaKatageri/hirokata (6 stars, last pushed 13d ago), licensed MIT. Its token cost is not measured: this kind of file is read by the harness, not the model. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other plugins, from other repositories
unslop
Make assistant output sound human. Strip AI-isms (sycophancy, stock vocab, hedging stacks, em-dash pileups), engineer burstiness, restore voice. Preserves code, URLs, and technical accuracy.
unslop
Humanize assistant output. Strip AI-isms, engineer burstiness, preserve technical accuracy.
simulator
Simulator.Company platform assistant. Exposes the Simulator REST API as an MCP server and provides skills for managing actors, forms, graph structures, and financial accounts.
courier
A bundle of Claude Code skills that ship work already on disk: commit-changes-to-git stages only what the user named and writes one Conventional Commits message without pushing, and submit-a-pull-request pushes the current branch, opens a pull request with a short prose body explaining what changed and why, and pings…
jamdesk
Build Jamdesk documentation sites with Claude. Complete reference for 50 MDX components, docs.json configuration, 8 navigation patterns, CLI commands, and writing standards.
adversaria
Critical thinking platform for decision analysis, proposal review, debate preparation, and due diligence. Three agents (advocate, adversary, judge) over a grounded argument graph, with a live transcript UI, appeal review, and 10 reasoning traditions.