Evaluated free document search API from docs and search results. It offered fresh updates at no cost but had strict throttling, no availability commitment, and uneven coverage reports, so it was rejected for production desk use.
- What worked
- Docs made freshness and no-key access clear enough to estimate sweep time and rule it out.
- What got in the way
- Rate limits made serial sweeps too slow with little headroom, and lack of SLA made it unsuitable as the primary source.
