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 pbvcity/supathink --skill st-initgit clone --depth 1 https://github.com/pbvcity/supathinkWrote 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/pbvcity/supathink/st-init)<a href="https://agentmods.dev/skills/pbvcity/supathink/st-init"><img src="https://agentmods.dev/badge/skills/pbvcity/supathink/st-init.svg" alt="Measured on agentmods" 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 | $0.00044 | $0.00649 |
| Opus 5 | $0.00022 | $0.00324 |
| Sonnet 5 | $0.00009 | $0.00130 |
| Haiku 4.5 | $0.00004 | $0.00065 |
Grade A, and why
st-init scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
用户在通过 supathink 插件壳请求安装或配置超限思考系统。本插件只提供命令入口;真正的 daemon 与宿主接线由 install.sh 完成。请带着用户一步步做,每步说明在做什么、征得同意后再执行(这是在用户机器上跑安装脚本,透明与确认是底线):
- 先检测是否已安装:
test -f ~/.supathink/bin/supathink && ~/.supathink/bin/supathink status。已装则跳到第 4 步做配置。 - 取得源码(询问用户偏好):
- 若本机已 clone 过 supathink 仓库,用其目录;否则
git clone https://github.com/pbvcity/supathink.git /tmp/supathink-src(或用户指定目录)。
- 若本机已 clone 过 supathink 仓库,用其目录;否则
- 运行安装:
cd <源码目录> && SUPATHINK_SKIP_CC_COMMANDS=1 ./install.sh(壳插件已提供命令,故跳过重复写入)。它会自动检测本机的 Claude Code / Codex / OpenClaw 并分别接线,daemon 落到 ~/.supathink。向用户复述它打印的"一次性动作"(Codex 需 /hooks 授信;OpenClaw 需重启网关)。 - 配置密钥(用 AskUserQuestion 或对话确认要配哪些;不要让用户把 key 贴进对话,给命令让其自己在终端执行):
- DeepSeek(强烈推荐,校验主力):
read -rs -p "DeepSeek API Key: " K && printf "\nSUPATHINK_DEEPSEEK_API_KEY=%s\n" "$K" >> ~/.supathink/env && unset K - 可选:GLM(合议中文席)、MiniMax(Judge)、Tavily(事实检索)——端点模板见 ~/.supathink/env,同法追加对应 SUPATHINK_*_API_KEY。
- DeepSeek(强烈推荐,校验主力):
- 选激活范围(AskUserQuestion):默认全关(仅显式命令触发)/ 某项目自动分档(该项目根写
.supathink.json{"auto":true})/ 全局自动(~/.supathink/env设SUPATHINK_AUTO=true)。解释默认关是推荐——日常零打扰。 - 收尾:
~/.supathink/bin/supathink status展示生效配置,并告诉用户:/st:slow <决策问题>触发一次深审试试,supathink log看过程,/st:win标记真实救场。
$ARGUMENTS
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.
- 7d ago First seen · 18 lines · 44 tokens per session scan A a0501485d957
st-init is a skill published in the GitHub repository pbvcity/supathink (5 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 44 tokens to every session and 649 once invoked, about $0.0002 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 2026-08-31.
Other skills, from other repositories
paper-verify-webchat
Use when the user asks to fact-check, verify, or score the citations/sources/references in a document from a chat (no terminal) — extracts citations with bundled code, fetches and judges each source with web tools, then scores through paper-verify's standard 100-point rubric via --from-evidence.
paper-verify
Use when the user asks to verify citations, audit source links, check blog/report/lecture references, or reduce a long bibliography to sources that need human review using the paper-verify CLI or MCP tools.
groundcheck
Verify factual claims against live web sources before asserting them. Use when about to state a fact you are not certain of (a date, number, name, statistic), or before publishing AI-generated text that makes factual claims. Returns a verdict (supported/refuted/unverified), a confidence score, and citations.
lenz-fact-check
Fact-check factual claims against independent web sources using the Lenz MCP server (the assessclaim / verifyclaim tools). Use this whenever the user wants to know whether something is actually true, asks you to verify or sanity-check a factual statement, wants a paragraph / article / blog post / document / dataset…
fact-check-social-media-posts
Verify claims in social media posts by checking evidence, evaluating source independence, and flagging rhetorical fallacies.
fact-check
Systematically fact-check AI-generated research reports and data-heavy documents. Trigger when the user asks to verify, fact-check, validate, or audit a report or document — especially those containing financial figures, market data, or claims sourced from web searches. Not for code review or general editing.