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 GiaSip/giasip-skills --skill giasip-researchgit clone --depth 1 https://github.com/GiaSip/giasip-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/giasip/giasip-skills/giasip-research)<a href="https://agentmods.dev/skills/giasip/giasip-skills/giasip-research"><img src="https://agentmods.dev/badge/skills/giasip/giasip-skills/giasip-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/skills/giasip/giasip-skills/giasip-research"><img src="https://agentmods.dev/badge/skills/giasip/giasip-skills/giasip-research.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00075 | $0.00452 |
| Opus 5 | $0.00037 | $0.00226 |
| Sonnet 5 | $0.00015 | $0.00090 |
| Haiku 4.5 | $0.00007 | $0.00045 |
Grade C, and why
giasip-research scanned grade C with 1 finding 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 4d 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.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
- Read and write only inside the output directory. Never `rm -rf`. This is a copy
97% identical to research — 2 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
What it actually says
✦ GiaSip · github.com/GiaSip/giasip-skills
GiaSip Research
Goal. Get the user accurate information that answers their question. Accuracy is a hard constraint; breadth is a means (you must gather widely to hit the accurate parts).
Method. Split the question into 2–3 complementary facets and run one sub-agent per facet in parallel. Sub-agents only gather, they do not conclude. Each sub-agent stops after at most 15 searches/fetches and returns compact findings without writing files. The main agent then writes a single report.md: answer the user's questions one by one, list every entity found with its URL, keep the body under 200 lines, and end with a separate "To verify" section.
Discipline (hard rules)
- Every fact carries a source URL. Prefer official / owner primary sources; label source type (primary / third-party / media).
- If it cannot be found, write "not found" — never fill from memory. Dates are absolute. GitHub stars and timestamps come from
gh apisnapshots, not search-result pages. - Read and write only inside the output directory. Never
rm -rf. - Before anything goes out under a real name, quotes a price, or states a legal/financial conclusion, run a verification pass on the claims.
Meta-rule. This skill only changes in response to a human comment made at the point of use (kept in FEEDBACK.md beside it). AI review feedback never edits it directly. Two controlled runs (2026-09-05/06) showed that growing this skill from 0 to 144 to 433 lines left precision flat, narrowed recall, and cost 6–13× more.
What ships with it
1 file 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.
- 4d ago Changed · -417 lines · -18 tokens per session scan A → C 412776cf73cc
- 12d ago First seen · 438 lines · 93 tokens per session scan A bfd3c2d2bf90
giasip-research is a skill published in the GitHub repository GiaSip/giasip-skills (12 stars, last pushed 3d ago), licensed MIT. It adds 75 tokens to every session and 452 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). It is 97% identical to research, differing in 2 lines, and is treated as a copy.
Other skills, from other repositories
prompt-master
Generates optimized prompts for AI tools. Activates only when the user explicitly asks to write, fix, improve, or adapt a prompt for a specific AI tool (LLM, Cursor, Midjourney, image AI, video AI, coding agents, etc.). Does not activate for general conversation, coding tasks, document writing, or other…
llm-wiki
Build and maintain a structured LLM-generated wiki for any codebase. Use when the user asks to analyze/understand/document a codebase, build a code wiki, create project documentation from source, or update an existing .llm-wiki. Triggers on phrases like "build wiki", "analyze this codebase", "document this project"…
pair-programming
A coding partner that writes code and then reviews that code for correctness, security, speed, readability, and reliability. It focuses on issues that can appear in AI-generated code, such as invented APIs or quietly changed requirements.
podcast-generator
A tool that turns Chinese text into a two-person conversational podcast using Volcengine’s Podcast AI service. It creates audio files in formats such as MP3, but requires a Volcengine application ID and access key.
product-manager
A product-management guide for turning software ideas into requirements, plans, and product documents. It also covers product strategy, growth, prioritisation, and competitor analysis.
deck-studio
A Chinese-language workflow for producing presentation decks, slide outlines, speaker notes, visual page drafts, and infographics.