@plurnk/plurnk-schemes-http-tavily
v1.19.1
Published
Tavily Extract page materialization for plurnk's http scheme — the third-party pluggability showcase.
Readme
@plurnk/plurnk-schemes-http-tavily
Tavily Extract page materialization for plurnk's http(s):// scheme — the
documented showcase of third-party pluggability. This package owns zero scheme
logic: it declares the http-materializer plugin family and implements the
materializer contract that @plurnk/plurnk-schemes-http selects.
Install and enable
The package ships with the plurnk service as a default dependency, so an ordinary service install already carries it. To install it beside a service that does not:
npm install @plurnk/plurnk-schemes-http-tavily# ~/.plurnk/.env, read by the service
PLURNK_SCHEMES_HTTP_MATERIALIZER=tavily-extract
TAVILY_API_KEY=...A generic public HTML READ then keeps the exact server source as the page's
body and lands the sanitized Tavily Markdown as its #readable channel; the
entry header records the materializer identity and Tavily's request/usage
evidence, and every READ of the page names #readable with its tokens. Without
the selection (or the key), the installed HTML projection produces #readable
instead, and the page reads the same way.
Configuration
| Variable | Contract |
| --- | --- |
| TAVILY_API_KEY | The Tavily API credential; absence makes the materializer ineligible |
| PLURNK_SCHEMES_HTTP_TAVILY_DEPTH | basic or advanced |
| PLURNK_SCHEMES_HTTP_TAVILY_TIMEOUT_MS | Positive request timeout in ms |
Contract
plurnk: { kind: "http-materializer", materializers: [{ id: "tavily-extract", module: "dist/materializer.js" }] }
exports one HttpMaterializer (eligible/extract) under the framework's
discovery, trust, and one-flat-id-namespace rules. See
SPEC.md for the exact outcome mapping.
Develop
npm install # links ../plurnk-service/plurnk-schemes-http for the plurnk-dev source
npm test