feat(sled): state, local and education procurement

[makegov-mark]

SDK support for Tango's new /api/sled/ namespace — state, local and education (SLED) procurement. Blocked: the endpoints are not live on the public API yet, and the vendored contract in this PR is ahead of what is deployed. Hold until the API ships them.

Python SDK equivalent: makegov/tango-python#62.

What's here

Method	Route
listSledOpportunities()	GET /api/sled/opportunities/
getSledOpportunity()	GET /api/sled/opportunities/{opportunity_id}/
listSledOpportunityRevisions()	GET /api/sled/opportunities/{opportunity_id}/revisions/
getSledCoverage()	GET /api/sled/opportunities/coverage/
listSledForecasts()	GET /api/sled/forecasts/
getSledForecast()	GET /api/sled/forecasts/{forecast_id}/

Plus iterateSledOpportunities / iterateSledForecasts and their IterableListMethod entries, the SledOpportunity / SledOpportunityRevision / SledForecast interfaces and six nested payload types, explicit shape schemas for all nine models, and five ShapeConfig constants. All 27 solicitation filters and all 13 forecast filters are typed options.

Four things a reviewer should check

The SDK must not synthesize status=open. The API defaults a list request with no status and no active to open-only. Making that explicit client-side breaks active: false, which is the complement of open and therefore has to reach the server as the only liveness param. Two tests pin both halves.

SLED_REVISIONS_MINIMAL omits changes. The per-field before/after requires a Small plan; naming it in a default shape would 403 every Free caller on a field they never asked to gate. changed_fields is in the default and available at every plan.

Support filters are filters only. platform, native_id and external_id are typed options and appear in no schema — a test asserts the registry has no such fields, so a caller can't shape on something the API will never return.

The shape tests walk nested expands, not just root leaves. This SDK has no parser.validate(); the existing parity test only checks root-level field names. client.sled.test.ts adds a walker that follows each expand through nestedModel, so attachments(size_bytes,…) and revisions(…,changes) are actually checked against SledAttachment / SledRevision rather than passing because nobody looked.

The fourteen nested sub-resource routes

Refreshing the contract described agencies_contracts_awarding, agencies_contracts_funding, contracts_subawards, entities_contracts, entities_idvs, entities_lcats, entities_otas, entities_otidvs, entities_subawards, idvs_awards, idvs_idvs, idvs_lcats, otidvs_awards and vehicles_orders for the first time. Unlike tango-python, this SDK treats an unmapped resource as a hard error, so the conformance gate went red with fourteen of them.

All fourteen already have SDK methods, so they are now mapped in RESOURCE_TO_METHOD. That converts fourteen errors into the honest finding: most of their options interfaces accept filters through an [key: string]: unknown index signature rather than declaring them, which the gate now reports as warnings. Declaring those is its own piece of work.

Their shape-coverage gaps are baselined instead, matching tango-python — each reuses its parent resource's model rather than carrying one of its own, so there is no model to map. None is SLED, and none is a regression.

Testing

• npx vitest run tests/ — 393 passed across 46 files; tests/unit/client.sled.test.ts adds 27.
• npm run typecheck — clean.
• npm run lint — clean.
• npm run check-conformance — 0 errors (was 14 after the contract refresh), 12 warnings.
• npm run check-shape-coverage — 0 new drift, 14 baselined.

npm run format reformats the whole repo, so the diff here is hand-scoped: three files (scripts/check-filter-shape-conformance.ts, scripts/check-shape-coverage.ts, tests/unit/config.shapes.parity.test.ts) were already prettier-dirty at main and carry only the added lines, not a reformat.

Docs

New State & Local (SLED) — Beta section in docs/API_REFERENCE.md, and an addition to docs/WEBHOOKS.md troubleshooting: an exclusion or DIBBS solicitation reaching its date fires nothing, because open/closed is derived at query time — but alerts.sled_opportunity.match does fire when a SLED solicitation closes, since its liveness is a stored field the API recomputes every fifteen minutes.

🤖 Generated with Claude Code

---

Update: the paid document-body leaf

Pushed a second commit adding attachments(extracted_text) — a SLED document body, served on the Small plan and above, from Tango API 4.25.1.

Three rules are documented wherever a caller will look, because each one is otherwise a support ticket:

• The leaf must be named. No default or suggested shape includes it, and attachments(*) does not carry it — the API resolves the body only for a caller who asked, so a default naming it would make every detail fetch pay for a document nobody wanted to read. A test pins the shapes against it.
• The key is absent, not null, whenever the text is not being served: below Small (where it is withheld and named in meta.upgrade_hints), on a contested document, or where the text could not be resolved.
• A contested document never returns text, at any plan — its stored bytes disagree with what the record advertised, so its extracted text is not reliably that record's content.

And the distinction worth stating plainly, because it reads backwards in both directions: searching document text is ungated (every plan, ids only, no fragment of the document) while reading a body is paid (Small and above, one record at a time). Buying the body does not change search.

Sequencing: the SLED surface itself is live on the production API. extracted_text needs 4.25.1, which is not deployed yet — so this PR is mergeable, but that one field will 400 until the API catches up. GET /api/version reports what a caller is talking to.