Help us improve
Share bugs, ideas, or general feedback.
From spicyregs
Queries regulatory data from Spicy Regs dataset using DuckDB against public Cloudflare R2 parquet bucket. Use for dockets, comments, documents, agency activity, filing timelines, cross-docket analysis.
npx claudepluginhub civictechdc/spicy-regs --plugin spicyregsHow this skill is triggered — by the user, by Claude, or both
Slash command
/spicyregs:spicyregsThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this skill to answer natural-language questions about regulatory data in this repo.
Checks regulatory feeds (US Federal Register, RSS) and reports new items filtered by materiality threshold. Detects coverage gaps against a source catalog.
Queries OpenData API datasets via REST for row fetching, filtering, sorting, aggregation, column inspection, and metadata retrieval. Use for data research, analysis, and pipelines on Parquet files.
Share bugs, ideas, or general feedback.
Use this skill to answer natural-language questions about regulatory data in this repo.
Prefer direct DuckDB queries against the public Cloudflare R2 bucket:
Keep this workflow provider-neutral.
uv, Python, and DuckDB.agents/openai.yaml file is only UI metadata for Codex-compatible surfaces. Non-OpenAI runtimes can ignore it and reuse the workflow in this skill plus references/provider-agnostic-prompt.md.uv run --script plugins/spicyregs/skills/spicyregs/scripts/query_spicy_regs.py --source r2 --list-sourcesuv run --script plugins/spicyregs/skills/spicyregs/scripts/query_spicy_regs.py --source r2 --describe docketsuv run --script plugins/spicyregs/skills/spicyregs/scripts/query_spicy_regs.py --source r2 --sql "<SQL>"uv run --script plugins/spicyregs/skills/spicyregs/scripts/find_duplicate_regulations.py --source r2 --limit 20uv run --script plugins/spicyregs/skills/spicyregs/scripts/query_spicy_regs.py --source local --list-sourcesreferences/data-layout.mdreferences/provider-agnostic-prompt.mdnotebooks/README.md--source local and inspect ./spicy-regs-data/ or the path they provide.sample-data/mirrulations/.--list-sources or --describe to confirm the remote tables and schema.LIKE match proves the claim.Use the helper script's --describe output for the exact schema. In this repo, the core datasets typically map to:
dockets: docket_id, agency_code, title, docket_type, modify_date, abstractdocuments: document_id, docket_id, agency_code, title, document_type, posted_date, modify_date, comment_start_date, comment_end_date, file_urlcomments: comment_id, docket_id, agency_code, title, comment, document_type, posted_date, modify_date, receive_date, attachments_jsoncomments_indexfeed_summaryagency_codedockets and documents on docket_idcomments by docket_id, date, or keywords in title and commentdocuments.comment_start_date and documents.comment_end_datenotebooks/query_data.ipynb and notebooks/cross_docket_analysis.ipynb when the question spans agencies or dockets.find_duplicate_regulations.py first, then verify promising clusters with targeted SQL against dockets or documents.Keep result sets small while exploring. Add LIMIT unless the user explicitly wants a full export.