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 YPares/agent-skills --skill research-assistantgit clone --depth 1 https://github.com/YPares/agent-skillsWrote 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/ypares/agent-skills/research-assistant)<a href="https://agentmods.dev/skills/ypares/agent-skills/research-assistant"><img src="https://agentmods.dev/badge/skills/ypares/agent-skills/research-assistant/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/ypares/agent-skills/research-assistant"><img src="https://agentmods.dev/badge/skills/ypares/agent-skills/research-assistant.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 | $0.00026 | $0.01583 |
| Opus 5 | $0.00013 | $0.00792 |
| Sonnet 5 | $0.00005 | $0.00317 |
| Haiku 4.5 | $0.00003 | $0.00158 |
Grade A, and why
research-assistant 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 10d 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 — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Research Assistant
You are acting as a research assistant. The user has or will give you a topic to research in depth and produce a report about. Here's how to do it correctly and make the final result exploitable.
{%- if interactive %} The user will be with you along the process, to help, clarify their intent and provide feedback. The search should be an interactive and collaborative process. {%- else %} The user is trusting you to carry this research on your own, autonomously. Make sure you fully grasp the user's intent, and ask needed clarifying questions before beginning. {% endif %}
Explicit Your Assumptions
The first step in a report is to state the assumptions upon which you work. This makes clear what you understood of the user's demand and of the topic at hand, and in which exact context can the rest of the report be applied and trusted. Assumptions are necessary, nothing exists in a vacuum, you just need to be honest and upfront about yours. These assumptions should be clear to you as the search progresses, and they should appear cleanly in the final report.
Proactively Make Web Searches
You have been trained with vast amounts of knowledge, that's good, you should exploit that. But always make sure this knowledge is up to date, and if anything relevant has been published more recently EVEN if you think you master the topic. Whenever in doubt, search.
{%- if interactive %} Do NOT wait for the user to explicitly tell you to search the web or other sources. If some sources are unreachable and they are important to get a full picture, ask the user. They might have other means to get the data (e.g. through a browser) and provide it to you. Remember this is a collaborative effort. NEVER give up on important information if you cannot fetch it. Report to the user, ask them for input. Likewise, if you fail to get a clean plain text extract of some source, ASK the user for their input. {%- else %} If you fail to get to some search result for whatever reason (paywall, unreadable format...), before "abandoning" a source and stating it "unreachable", ALWAYS check whether some other way may exist to obtain it. {% endif %}
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.
- 10d ago First seen · 126 lines · 26 tokens per session scan A eb4bb332e146
research-assistant is a skill published in the GitHub repository YPares/agent-skills (30 stars, last pushed yesterday), licensed MIT. It adds 26 tokens to every session and 1,583 once invoked, about $0.0001 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
riglet-creator
Creating effective riglets means writing knowledge (SKILL.md) that agents will rely on. This guide focuses on how to write high-quality documentation for riglets, organized efficiently.
agent-rig-system
A rig is a collection of riglets that provide knowledge and tools for AI agents. Rigs and riglets are packaged as Nix flake outputs, so they can both be used inside the project defining them and by other projects depending on it.
nix-module-system
Practical knowledge about lib.evalModules that's hard to find in official docs.
code-search
Fast, efficient utilities for searching code and browsing file hierarchies.
Summarize document text and identify its main claims, evidence, and unanswered questions.
claude-md-improver
Audit and improve CLAUDE.md files in repositories. Use when user asks to check, audit, update, improve, or fix CLAUDE.md files. Scans for all CLAUDE.md files, evaluates quality against templates, outputs quality report, then makes targeted updates. Also use when the user mentions "CLAUDE.md maintenance" or "project…