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.
npx skills add WenyuChiou/research-hub --skill research-hub-multi-aigit clone --depth 1 https://github.com/WenyuChiou/research-hubWrote 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/skills/wenyuchiou/research-hub/research-hub-multi-ai)<a href="https://agentmods.dev/skills/wenyuchiou/research-hub/research-hub-multi-ai"><img src="https://agentmods.dev/badge/skills/wenyuchiou/research-hub/research-hub-multi-ai/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/skills/wenyuchiou/research-hub/research-hub-multi-ai"><img src="https://agentmods.dev/badge/skills/wenyuchiou/research-hub/research-hub-multi-ai.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00100 | $0.01006 |
| Opus 5 | $0.00050 | $0.00503 |
| Sonnet 5 | $0.00020 | $0.00201 |
| Haiku 4.5 | $0.00010 | $0.00101 |
Grade A, and why
research-hub-multi-ai 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 9d 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.
How it starts
The opening of the file, as written. The whole thing — 108 lines — stays where its author put it; the contents beside it link to each section on GitHub.
research-hub Multi-AI Router
This is a planning router. It writes a coordination plan and per-leaf briefs; the supported leaves execute. The current leaf set is:
codex-delegate: implementation-heavy or repetitive work with a structuredresult.jsoncontract.antigravity-delegate: narrowly scoped, non-honesty-critical mechanical work with anagy_result_*.mdcontract.
The archived Gemini delegate is not a supported leaf. Long-context reading, CJK drafting, scientific judgment, governance, and final review stay with the primary model unless the operator has separately verified another adapter.
When to invoke
Invoke only when a single round needs two or more delegates and at least one task operates on research-hub artifacts. Examples:
- Codex scaffolds tests while Antigravity applies an already-approved metadata transform to a disjoint fixture subtree.
- Two independent Codex tasks prepare separate adapters, followed by primary model reconciliation.
- Antigravity performs a bounded transcription, then Codex consumes the verified artifact to generate a deterministic report.
For one delegate, use its leaf skill directly. For generic, non-research
multi-agent work, use agent-collab-workspace:agent-task-splitter.
Do not invoke
- One Codex or Antigravity task.
- A single
research-hub auto --llm-cli ...run. - Translation, long-form CJK drafting, literature judgment, security, governance, or final review. Keep those with the primary model.
- Work whose delegates would edit overlapping files in parallel.
Prerequisite and health checks
Before writing a plan:
- Verify every named leaf skill and wrapper exists.
- Run the leaf's documented preflight (
codex --versionoragy --version). - If the plan calls
research-hub, runresearch-hub doctor. - Record an unavailable leaf as a blocker. Never silently substitute an unverified executor.
If the research-hub CLI is missing, do not emit commands that call it. Plans that use only healthy delegate wrappers may proceed.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 9d ago Changed · -9 lines · -54 tokens per session 3431ba5677b9
- 13d ago First seen · 117 lines · 154 tokens per session scan A 5f79766aea28
research-hub-multi-ai is a skill published in the GitHub repository WenyuChiou/research-hub (54 stars, last pushed 4d ago), licensed MIT. It adds 100 tokens to every session and 1,006 once invoked, about $0.0005 per session on Opus 5. 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 2026-08-30.
Other skills, from other repositories
migrate-v0-to-v1
Migrate a Scraps wiki from v0 to v1. Use this whenever the user asks to upgrade Scraps, migrate v0 wiki syntax/config, convert implicit tags to explicit #[[tag]], move from scraps-writer to scraps, or audit a wiki for v1 readiness. This skill is designed for one-pass LLM-led migrations that discover the latest v0 and…
remem
Use when the user asks Codex to recall prior project context, save durable decisions or bug fixes, inspect remem memory health, or activate remem automatic memory hooks from the Codex plugin.
ingest
Create a new scrap from a source (prompt, URL, or arbitrary markdown), update cross-links in related scraps, and run a sanity check. Use this when the user wants to add a scrap, summarize an article, save synthesized content to the wiki, or file back a Q&A result.
query
Answer questions about the wiki by searching, reading, and synthesizing relevant scraps with [[Title]] citations. Use this when the user wants to query the wiki, ask what they have written about a topic, compare scraps, find related notes, or pull a synthesized overview from existing scraps.
planning-with-scratchpad
Use when user explicitly requests planning with a scratchpad, or asks for persistent tracking of a complex task that the human can browse visually. Backs planning files with the scratch CLI so a pad's files are registered and viewable.
scratch
Organize knowledge from an agent session into a scratchpad — a folder of files plus a scratchpad.json manifest. Use when a task generates notes, snippets, command output, diagrams, or artifacts worth keeping together for a session/activity.