Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/Rheinmir/setupnpx agentmods add skills/rheinmir/setup/dym-scientific-agent-skills-gate-trong-ci-cua-banWrote 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/rheinmir/setup/dym-scientific-agent-skills-gate-trong-ci-cua-ban)<a href="https://agentmods.dev/skills/rheinmir/setup/dym-scientific-agent-skills-gate-trong-ci-cua-ban"><img src="https://agentmods.dev/badge/skills/rheinmir/setup/dym-scientific-agent-skills-gate-trong-ci-cua-ban/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/rheinmir/setup/dym-scientific-agent-skills-gate-trong-ci-cua-ban"><img src="https://agentmods.dev/badge/skills/rheinmir/setup/dym-scientific-agent-skills-gate-trong-ci-cua-ban.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 | $0.00079 | $0.02014 |
| Opus 5.5 | $0.00032 | $0.00806 |
| Sonnet 5.5 | $0.00016 | $0.00403 |
| Haiku 4.5 | $0.00008 | $0.00201 |
Grade A, and why
dym-scientific-agent-skills-gate-trong-ci-cua-ban 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 yesterday.
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 — 153 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Skill: dym-scientific-agent-skills-gate-trong-ci-cua-ban — Đưa skill vào CI của DỰ ÁN BẠN
Đây là ví dụ viết mới, tối thiểu, cho người dùng. Nó không phải bản sao
.github/workflows/*.ymlcủa repo skill — bốn workflow trong đó (skill-tests,skill-spec-validation,pr-skill-scan,security-scan) tồn tại để kiểm chính repo skill, cần secret riêng của họ, và không có ý nghĩa gì trong dự án của bạn. Nếu bạn đóng góp cho repo gốc thì xem 08.
Vì sao dùng: skill là chỉ dẫn agent tuân theo. Cắm chúng vào CI để (a) skill được commit đúng như lúc bạn thẩm định, (b) upstream thay đổi thì CI báo thay vì agent lặng lẽ đổi hành vi. Sinh ra cái gì: một job CI đỏ khi skill đã cài trôi khỏi bản đã duyệt, hoặc khi một skill không còn hợp lệ theo spec.
Quyết định trước: commit skill hay cài lúc chạy CI
| Commit vào repo | Cài trong CI | |
|---|---|---|
| Tái lập | tuyệt đối — đúng byte bạn đã đọc | phụ thuộc tag/SHA remote |
| Review | diff hiện trong PR, người soát thấy | thay đổi vô hình |
| Kích thước repo | tăng (một skill ~4 KB–1.2 MB) | không |
| Khuyến nghị | ✅ mặc định nên chọn cái này | chỉ khi repo bạn cấm vendor |
Commit skill vào repo là mặc định đúng: bạn đã bỏ công thẩm định ở 02, commit là cách duy nhất giữ được kết quả thẩm định đó.
Gate 1 — Phát hiện drift so với upstream (đã chạy thật)
gh skill install ghi github-tree-sha vào frontmatter lúc cài. So nó với tree SHA remote hiện tại là phát hiện được upstream đã sửa skill.
Lưu thành ci/check-skill-drift.sh:
#!/usr/bin/env bash
# Đỏ nếu một skill đã cài trôi khỏi tree SHA ghi lúc cài.
set -euo pipefail
status=0
for skill in "$@"; do
f=".claude/skills/$skill/SKILL.md"
[ -f "$f" ] || { echo "MISSING $skill"; status=1; continue; }
repo=$(awk '/github-repo:/{print $2; exit}' "$f" | sed 's#https://github.com/##')
ref=$(awk '/github-ref:/{print $2; exit}' "$f")
path=$(awk '/github-path:/{print $2; exit}' "$f")
pinned=$(awk '/github-tree-sha:/{print $2; exit}' "$f")
parent=$(dirname "$path"); name=$(basename "$path")
remote=$(gh api "repos/$repo/git/trees/${ref#refs/tags/}:$parent" \
--jq ".tree[] | select(.path==\"$name\") | .sha")
if [ "$pinned" = "$remote" ]; then
echo "OK $skill $pinned"
else
echo "DRIFTED $skill local=$pinned remote=$remote"; status=1
fi
done
exit $status
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.
- yesterday First seen · 153 lines · 79 tokens per session scan A 089a34d1a7ac
dym-scientific-agent-skills-gate-trong-ci-cua-ban is a skill published in the GitHub repository Rheinmir/setup (11 stars, last pushed 20d ago), licensed MIT. It adds 79 tokens to every session and 2,014 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-11-07.
dym-scientific-agent-skills-gate-trong-ci-cua-ban compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| dym-scientific-agent-skills-gate-trong-ci-cua-banRheinmir/setup | Skill | 11 | 20d ago | MIT | A | 79 | Claude Code |
| chinese-git-workflowjnMetaCode/superpowers-zh | Skill | 8.5k | 12d ago | MIT | A | 69 | Claude Code |
| atmos-profilescloudposse/atmos | Skill | 1.4k | today | Apache-2.0 | A | 35 | Claude Code |
| webhook-managementhoangsonww/Claude-Code-Agent-Monitor | Skill | 1.2k | yesterday | MIT | A | 56 | Claude Code, Codex |
| monorepo-managementwshobson/agents | Skill | 42k | 2d ago | MIT | C | 54 | Claude Code |
| managing-github-actions-secretsPostHog/posthog-foss | Skill | 728 | today | MIT | A | 67 | Claude Code, Codex |
| postman-newman-automationLambdaTest/agent-skills | Skill | 391 | 2d ago | MIT | A | 156 | Claude Code, Codex |