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 unbias38/my-claude-skills --skill interpreting-stock-moodgit clone --depth 1 https://github.com/unbias38/my-claude-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/unbias38/my-claude-skills/interpreting-stock-mood)<a href="https://agentmods.dev/skills/unbias38/my-claude-skills/interpreting-stock-mood"><img src="https://agentmods.dev/badge/skills/unbias38/my-claude-skills/interpreting-stock-mood/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/unbias38/my-claude-skills/interpreting-stock-mood"><img src="https://agentmods.dev/badge/skills/unbias38/my-claude-skills/interpreting-stock-mood.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 | $0.00199 | $0.08559 |
| Opus 5 | $0.00100 | $0.04279 |
| Sonnet 5 | $0.00040 | $0.01712 |
| Haiku 4.5 | $0.00020 | $0.00856 |
Grade C, and why
interpreting-stock-mood 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 12d 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.
Hidden instructionshighPrompt injection
Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.
> 🧑🤝🧑 散戶熱度:0%(10 頁 169 篇未中)→ 冷清 問題:使用者以為散戶真的不討論,但其實是 PTT 連線失敗。 ✅ **新版(透明)**: > 🧑🤝🧑 散戶熱度:⚠️ 抓取失敗(未納入判讀) 或如果是真的搜了沒中: > 🧑🤝🧑 散戶熱度:0%(搜 [2603, 長榮] 在 169 篇未中)→ 冷清 ### Step 6:警訊判斷 若技術特徵符合下列任一警訊模式,**必須**在「建議行動」中以 How it starts
The opening of the file, as written. The whole thing — 450 lines — stays where its author put it; the contents beside it link to each section on GitHub.
股票溝通師 / 解讀股票心情(Interpreting Stock Mood)
從股價、新聞、社群熱度三個資料來源萃取客觀特徵,結合技術分析常識,用擬人化獨白搭配寫實技術解讀,產出市場情緒推測與行動建議。
重要前提:這是「解讀」不是「預測」
科學上沒有任何方法能準確預測未來股價。本技能的輸出皆為基於技術指標與市場資訊的合理解讀,並會在輸出中明確標註信心度。禁止用「他說明天會漲」這種斷言語氣,應使用「依據近期技術特徵推測,這檔股票目前較可能對應到 xxx 市場情緒」的描述語氣。
所有輸出必須附帶免責聲明:「⚠️ 本分析僅為技術面解讀與娛樂用途,非投資建議。投資決策請自行評估並承擔風險。」
觸發判斷
當使用者同時提供(或明顯打算提供)以下兩項時觸發:
- 股票代號(台股如「2330」「2330.TW」,美股如「AAPL」「TSLA」,ETF 如「0050」「QQQ」)
- 想詢問的事項(例如「他最近怎麼了」「他是不是要噴了」「我該不該買」)
若使用者只給其中一項,主動詢問缺少的部分,不要自行猜測。
工作流程
Step 1:確認兩項輸入齊全
- 股票代號(自動補上 .TW 後綴判斷台股,例:2330 → 2330.TW;4 位數字一律視為台股。上市 .TW 抓不到時腳本會自動改試上櫃 .TWO,不用手動指定)
- 提問(例:「他最近是不是不太理我?」「我該不該加碼?」)
Step 2:判斷提問是否模糊,需要反問
如果使用者的提問沒有同時包含以下四個面向中的至少兩個,視為「模糊提問」,先反問 3-4 題再進行分析:
| 面向 | 例子 |
|---|---|
| 持倉狀態 | 已持有 / 想買 / 想賣 / 觀望 |
| 方向 | 看多 / 看空 / 不確定 |
| 金額或部位 | 已投入多少 / 想投入多少 / 佔資產比例 |
| 時間範圍 | 短線(< 1 個月)/ 中線(1-6 個月)/ 長線(> 6 個月) |
模糊提問範例:
- 「台積電最近都不太理我」
- 「他是不是在生氣」
- 「他是不是要噴了」
反問格式(Markdown):
我可以幫你跟 {代號} 溝通看看 🎧,不過為了讓分析更貼合你的處境,先問你 4 件事:
1. 📦 **你目前的持倉狀態**?
- (a) 已經持有 (b) 還沒買,正在考慮 (c) 想賣,正在考慮 (d) 純粹好奇
2. 🧭 **你的方向偏好**?
- (a) 看好(想長期持有 / 加碼) (b) 看壞(想停損 / 放空) (c) 不確定 (d) 跟著市場走就好
3. 💰 **金額或部位規模**(隨意,可只說大概)?
- 例:已投入 10 萬 / 想投入單筆 / 佔資產 20% / 不重要
4. ⏳ **時間範圍**?
- (a) 短線(一個月內) (b) 中線(半年內) (c) 長線(半年以上) (d) 沒概念
你回答後我再進行完整分析(後台會自動抓股價、新聞、社群熱度)。
清楚提問範例(直接進 Step 3,不需反問):
- 「我手上 0050 持有 3 年,現在想加碼,他現在是好時機嗎?」
- 「我想短線買 2454,他這兩週狀況怎麼樣?」
Step 3:跑技術指標 + 多維熱度提取腳本
執行 scripts/analyze_stock.py:
python "<SKILL_DIR>/scripts/analyze_stock.py" "<股票代號>" "<別名清單>"
📌
<SKILL_DIR>是什麼:就是這個 SKILL.md 所在的資料夾絕對路徑(你讀到這份 SKILL.md 時就已經知道它在哪)。 跨平台寫法:路徑請整段用雙引號包起來(Windows 路徑常含空格與反斜線;macOS/Linux 路徑可能含中文),用一行寫完,不要用\(bash)或`(PowerShell)做行繼續 —— 直接呼叫 Bash/PowerShell tool 一行交出去最穩。
第二個參數(別名清單)— 很重要,影響 PTT 搜尋命中率
支援兩種格式:
- 單一中文名:
聯發科 - 逗號分隔的多個別名(推薦):
聯發科,聯發,MTK,MediaTek
為什麼要傳多別名? PTT 股板的標題對同一支股票會用不同講法 —— 「聯發科」「MTK」「MediaTek」都有人寫。多傳幾個別名命中率高很多,避免出現「PTT 冷清(其實只是搜不到)」的誤判。
What ships with it
10 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.
- .gitignore 286 B
- assets/fonts/NOTICE.md 656 B
- assets/fonts/NotoSansTC.ttf 11662 KB
- examples/usage_example.md 14 KB
- README.md 8.1 KB
- references/aliases.md 4.5 KB
- references/asset_traits.md 13 KB
- references/technical_emotion_mapping.md 11 KB
- scripts/analyze_stock.py 59 KB runs code
- scripts/generate_report.py 42 KB runs code
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.
- 12d ago First seen · 450 lines · 199 tokens per session scan C 20f28a56799e
interpreting-stock-mood is a skill published in the GitHub repository unbias38/my-claude-skills (2 stars, last pushed 17d ago), licensed MIT. It adds 199 tokens to every session and 8,559 once invoked, about $0.0010 per session on Opus 5. A static security scan graded it C with 1 finding (hidden instructions). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
agb-bei-kreditvertraegen-verbraucherdarlehen
Für AGB bei Kreditverträgen Verbraucherdarlehen: ordnet Norm, Beweislast und Gegenargument; Ergebnis: Prüfprodukt mit Risiko und nächstem Schritt.
agb-in-kapitalanlagen-effektenhandel
Für AGB in Kapitalanlagen Effektenhandel: ordnet Norm, Beweislast und Gegenargument; Ergebnis: Prüfprodukt mit Risiko und nächstem Schritt.
app-planner
Guides you through comprehensive iOS/Swift app planning and analysis. Use for new apps (concept to architecture) or existing apps (audit current state, plan improvements, evaluate tech stack). Covers product planning, technical decisions, UI/UX design, and distribution strategy.
amazon-returns-recovery
Suede-affiliated Amazon money-recovery audit for restocking fees, short or denied refunds, and Amazon-billed subscriptions. Use when the user mentions a restocking fee, a short or denied Amazon refund, or a forgotten Prime Video Channel, Audible, Kindle Unlimited, or Prime charge, or asks whether Amazon still owes or…
subscription-recovery
Suede-owned recovery discipline for recurring charges billed outside Amazon: App Store, Google Play, PayPal, direct-bill streaming, gyms, news, and SaaS. Use when the user wants to find, audit, cancel, or dispute a subscription they may have forgotten, is being charged for twice, or no longer uses. Every cancellation…
opendeviation-eval-metrics
Machine-readable reference + computation scripts for state-of-the-art metrics evaluating open deviation bar (ODB, brim-to-brim price-based sampling) data.