菜单
Track legislation, regulatory actions, and legal developments affecting prediction markets, crypto, and AI agents — triaged by stage × impact for decision-ready output
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/ThomasMoreAI/legal-skills-open --skill regulatory-monitor命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Use when retrieving arbitration institutional rules (ICC, LCIA, SCC, SIAC, HKIAC, VIAC, МКАС/МАК при ТПП України, UNCITRAL) — fetching current version, verifying redaction applicable to the date of arbitration agreement, constructing URLs for official rule texts
Use when choosing the Polish legal regime for letters, requests, applications, complaints, petitions, public-information requests, KPA filings, PPSA complaints, RODO access requests, registry extracts, court-file access, tax/ZUS/cudzoziemcy/USC procedures, or professional lawyer letters. Prevents mixing UDIP, KPA, PPSA, RODO, registry, special-procedure, and advocate/radca letter regimes.
Use when preparing applications for recognition and enforcement of foreign arbitral awards in Poland, applications for setting aside arbitral awards under KPC art. 1205–1211, or opposing such applications — mapping Article V of the 1958 New York Convention to art. 1214–1215 of the Polish KPC, identifying grounds for refusal, structuring public policy arguments
基于 SOC 职业分类
正在显示 SKILL.md
| name | regulatory-monitor |
| title | Thesis |
| description | Track legislation, regulatory actions, and legal developments affecting prediction markets, crypto, and AI agents — triaged by stage × impact for decision-ready output |
| author | aaronjmars |
| author_url | https://github.com/aaronjmars/aeon/tree/main/skills/reg-monitor |
| license | MIT |
| version | 0.1.0 |
| execution_mode | open |
| jurisdiction | us |
| practice | regulatory |
| language | en |
| tags | ["crypto","research"] |
${var} — Specific regulatory topic to narrow focus (e.g.
prediction-markets,stablecoins,ai-agents). If empty, scans all tracked domains.
Read memory/MEMORY.md for context on current positions and interests.
Read the last 2 days of memory/logs/ entries for this skill to avoid repeating items.
Read memory/topics/reg-monitor-seen.md (create if missing) for persistent URL dedup across all prior runs.
Original skill produced news-shaped output (HIGH PRIORITY / NOTABLE / WATCH LIST) with soft ranking and no deadlines. Reg-monitor's real value is decisions — file a comment, prepare for a ban, short a market, brief counsel. Rank items by stage × impact, surface deadlines, drop filler buckets, and distinguish "quiet week" from "monitor failure" via source-status observability.
Set today=$(date -u +%F), year=$(date +%Y), since=$(date -u -d '7 days ago' +%F).
Fetch from each source below. Track which succeed/fail — you'll emit a source_status footer. Treat each as independent; any single failure is non-fatal.
A. Federal Register (authoritative US rulemakings, notices, enforcement) — public, no auth:
WebFetch: https://www.federalregister.gov/api/v1/documents.json?conditions[publication_date][gte]=${since}&conditions[term]=prediction+market+OR+digital+asset+OR+stablecoin+OR+cryptocurrency+OR+AI+agent&per_page=40&order=newest
Extract per item: title, publication_date, type (Rule / Proposed Rule / Notice), agencies, document_number, html_url, comments_close_on (critical — drives ACT bucket), abstract.
B. SEC press releases (RSS) — authoritative enforcement + guidance:
WebFetch: https://www.sec.gov/news/pressreleases.rss
Keep only items published on/after ${since}. Match titles against keywords: crypto, digital asset, stablecoin, token, DeFi, prediction, event contract, AI.
C. CFTC press releases (RSS) — authoritative prediction market + derivatives actions:
WebFetch: https://www.cftc.gov/RSS/RSSPR/rsspr.xml
Keep items on/after ${since}. CFTC runs prediction-market rulemaking — all CFTC items are in-scope by default.
D. WebSearch (gap-filler for state bills, court rulings, international) — run these with the current year injected:
"prediction market" (bill OR legislation OR ban OR ruling) ${year}"Polymarket" OR "Kalshi" (CFTC OR state OR ruling OR investigation) ${year}stablecoin (legislation OR regulation OR enforcement) ${year}"AI agent" (regulation OR liability OR legislation) ${year}ESMA OR MiCA (enforcement OR guidance) ${year}If ${var} is set, replace the above with 3 targeted searches for that specific topic (include ${year} and "bill OR ruling OR enforcement").
E. Congress.gov (optional, skip if no API key) — if CONGRESS_GOV_API_KEY env var is set, WebFetch bill search for crypto/prediction-market keywords. Otherwise skip silently and rely on Federal Register + WebSearch to catch bill activity.
Read memory/topics/reg-monitor-seen.md (format: one URL per line, oldest first). Drop any candidate whose URL already appears. Also drop candidates whose title substring-matches an entry from the last 2 days of memory/logs/.
If the seen-file exceeds 500 lines, keep only the most recent 500 before appending new items at end.
Discard:
Keep only: bills introduced/advancing, agency rulemakings, enforcement actions, court rulings, official guidance, international regulatory coordination.
Compute triage_score = impact × stage:
impact (domain relevance to tracked interests):
stage (how close to binding):
Jurisdiction multiplier — apply AFTER impact × stage:
Cap at 8 items total across all buckets. Sort within each bucket by score descending.
For every ACT item and top 2 WATCH items, WebFetch the source URL and extract:
Keep under 4000 chars. If all buckets empty after filtering: skip notify, write REG_MONITOR_OK to the log, end.
*Reg Monitor — ${today}*
{1-line thesis: e.g. "3 items need a decision this week" or "quiet week, 2 to watch"}
*ACT*
• [Title](url) — jurisdiction · stage
what changed. who's affected.
⏰ {deadline or effective date}
📊 prediction market angle: {if any}
*WATCH*
• [Title](url) — jurisdiction · stage
one-sentence verdict.
*CONTEXT* (only if included)
• [Title](url) — one-line summary.
—
sources: federal-register={ok|fail} · sec={ok|fail} · cftc={ok|fail} · websearch={ok|fail} · congress={ok|fail|skip}
{n} developments scored, {n_act} act / {n_watch} watch / {n_context} context
If all sources failed, send a different notification: *Reg Monitor — ${today}* ⚠️ all sources failed, see log and write REG_MONITOR_ERROR to the log. Do not emit an empty digest.
If some sources failed (partial data), still send the digest but flag it in the thesis line: ⚠️ partial — {failed sources} down. Log REG_MONITOR_DEGRADED.
Append every URL included in the notification to memory/topics/reg-monitor-seen.md, one per line.
memory/logs/${today}.md### reg-monitor
- Status: {REG_MONITOR_OK | REG_MONITOR_DEGRADED | REG_MONITOR_ERROR | sent}
- Items: {n_act} act / {n_watch} watch / {n_context} context
- Sources: fr={ok|fail} sec={ok|fail} cftc={ok|fail} ws={ok|fail} cg={ok|fail|skip}
- Top item: {title, score, stage}
- Deadlines tracked: {any comment-period or vote dates in the next 14 days}
Include: bills introduced/advancing; agency rulemakings (NPRM/ANPRM/final rule); enforcement actions (fines, C&D, consent orders); court rulings affecting in-scope domains; official guidance/advisory letters; international coordination statements tied to a concrete document (FATF, G7, Basel).
Exclude: opinion pieces; think-tank reports (unless cited in a bill/docket); industry lobbying with no bill attached; academic papers; speculation without a document; press releases announcing future press releases.
The sandbox blocks some outbound curl. Use WebFetch for every URL fetch (it's the Claude built-in and bypasses the sandbox). Federal Register, SEC RSS, CFTC RSS, and WebSearch all work without env-var auth. Congress.gov API requires CONGRESS_GOV_API_KEY — skip silently if unset.
CONGRESS_GOV_API_KEY (optional) — enables Congress.gov bill search step E. Skill degrades gracefully without it.REG_MONITOR_OK and stay silent.memory/topics/reg-monitor-seen.md.