pull-data
Skill2 starsChanged 5 months agoMITView on GitHub
Use when the user wants to fetch/download Greek public-procurement data (Διαύγεια award decisions, ΚΗΜΔΗΣ contracts, TED EU tenders) into local files for investigation. Triggers on "pull", "download", "get the data", "fetch contracts/awards", a date range, or an authority/supplier to look at.
In this repository4 other files it gives its agents.
- build-corpusSkill
- detect-red-flagsSkill
- investigateSkill
- write-briefSkill
The file
Files
Pull Greek procurement open data
You fetch real, public data into local JSONL files the user owns. No account, no key (except
ΓΕΜΗ). The scripts live in lib/. Run them from the repo root.
Which source to pull
- ΚΗΜΔΗΣ (contracts) — the default and best. ~100% have a supplier ΑΦΜ, amount, CPV,
and the signing official. Use for almost everything.
→node lib/kimdis.mjs --days=14 # last 14 days, national node lib/kimdis.mjs --from=2026-05-01 --to=2026-05-31data/contracts.jsonl - Διαύγεια (awards) — every government decision, incl. tiny ones; ~half carry a structured
amount. Use for breadth or when ΚΗΜΔΗΣ misses something.
→node lib/diavgeia.mjs --target=4000 node lib/orgs.mjs # then resolve org-ids → authority namesdata/awards.jsonl - TED — EU above-threshold notices/awards for Greek buyers (the big contracts):
node lib/ted.mjs --type=can-standard
Then
- Resolve supplier VATs to legal names (companies):
node lib/vies.mjs --from=data/contracts.jsonl - Hand off to the detect-red-flags skill (find splitting/fragmentation) or the investigate skill (follow a supplier/authority/official).
Guidance
- Pulling is throttled to respect the public APIs — for big ranges, narrow the dates or raise
--maxpages. Tell the user roughly how many records came back. - Bigger samples = more (and stronger) leads. A couple of weeks of ΚΗΜΔΗΣ is a good start.
- These are PUBLIC sources; the user is just reading open data. Always keep the
leads-not-accusations framing (see
docs/METHODOLOGY.md).
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.