@pipeworx/jmpr

Connect: https://gateway.pipeworx.io/jmpr/mcp · Install: one-click buttons

Tools: 3

Acceptable Daily Intakes (ADI) and Acute Reference Doses (ARfD) set by the Joint FAO/WHO Meeting on Pesticide Residues — the toxicological assessment that every Codex pesticide limit is derived from.

Tools

The three-pack picture

The gaps between them carry meaning. Chlorpyrifos is the worked example: Codex revoked its residue limits, so codex-mrl has no record of it at all — while JMPR still holds the evaluation, and the 2024 entry says why: the toxicological database was insufficient to establish health-based guidance values. Neither pack alone tells that story; codex-mrl would report a bare absence.

Coverage

382 pesticides. Evaluations run from the 1960s to the present, newest first.

Auth

None. Keyless, no registration, no quota.

Data source

https://apps.who.int/pesticide-residues-jmpr-database/ — server-rendered HTML, no API. Two shapes: /Home/Range/All lists every pesticide, /pesticide?name=NAME is one record. Names are the identifier — there are no numeric ids, so a name that doesn’t match exactly renders an empty shell rather than a 404, which is why both detail lookups resolve against the catalogue first.

Gotchas

1. A blank ADI is a finding, not a gap. The evaluation table has an ADI column that is often empty, and dropping those rows or treating them as “no data” inverts the meaning. Chlorpyrifos 2024 has no ADI because the meeting judged the toxicological data insufficient — and the reason sits in the Comments cell of that same row. When a pesticide has been evaluated but carries no ADI anywhere, jmpr_chemical says so in note.

2. The newest evaluation is not always the newest ADI. A re-evaluation can decline to set a value while the previous one still stands, so current_adi / current_adi_year are tracked separately from latest_evaluation_year. For chlorpyrifos those are 2004 and 2024 respectively.

3. Absence from this database means never evaluated. That is different from “evaluated and found unsafe” — the no_match hints say which, because an agent reading a bare empty result will assume the stronger claim.

4. Parent compounds and variants are separate records with different ADIs (CHLORPYRIFOS vs CHLORPYRIFOS-METHYL). A fragment that matches several returns all of them flagged ambiguous rather than picking one; an exact name still resolves outright.

5. Document links are mixed relative and absolute — ./Document/297 on the WHO host, or a full FAO URL. Both are resolved to absolute.

Tools

• jmpr_search — Find pesticides evaluated by the FAO/WHO Joint Meeting on Pesticide Residues (JMPR) by name or partial name. Returns matching pesticide names, which are the identifier used to fetch a full safety eval
• jmpr_chemical — Get the full JMPR safety evaluation for one pesticide: its Acceptable Daily Intake (ADI) and Acute Reference Dose (ARfD) in mg/kg body weight, the year each was set, the meeting’s conclusions, its CAS
• jmpr_list_pesticides — List every pesticide the FAO/WHO Joint Meeting on Pesticide Residues has evaluated. Use to survey coverage or to check whether a substance has ever been assessed at all — absence means JMPR has not ev