Read-only JSON endpoints.

Full feed: every tool we track, with editorial fields, plans, prices, limits, included and missing features, source URLs and capture timestamps. Index-style.

Top-level keys

feed_version, generated_at, tools[], per_tool_endpoints[]

Shape

{ feed_version, generated_at, count, tools: Tool[],
  per_tool_endpoints: string[] }

Per-tool feed: a single Tool with the full editorial + plan payload. One endpoint per slug; the index feed lists them all in `per_tool_endpoints`.

Top-level keys

feed_version, generated_at, tool

Shape

{ feed_version, generated_at, tool: Tool }

Per-category feed: all tools whose primary or secondary category matches the slug, ranked the same way as `/best/<category>`. One endpoint per category.

Top-level keys

feed_version, generated_at, category, tools[]

Shape

{ feed_version, generated_at, category, tools: Tool[] }

Per-category pricing-comparison feed: cost-only slice (free tier, entry paid plan, mid-tier paid plan) per tool. Mirrors the HTML view at /pricing/<category>. One endpoint per category with 2+ tools. Designed for AI search engines to cite when answering 'compare X pricing' queries.

Top-level keys

feed_version, generated_at, category, page_url, tools[]

Shape

{ feed_version, generated_at, category, page_url,
  tools: { slug, name, plans: { free, entry_paid, mid_tier } }[] }

Pricing-change events: each entry is a captured anchor-price change from the weekly scrape. Empty until the cron has run with real changes; the URL is stable so consumers can subscribe-by-polling.

Top-level keys

feed_version, generated_at, count, events[]

Shape

{ feed_version, generated_at, count, events: PriceChange[] }

Coverage gaps: tools referenced by something we DO track (via `alternatives`) but not yet seeded into the spine. Mirrors the data on /coverage. Sorted by reference count, descending.

Top-level keys

feed_version, generated_at, tools_tracked, gap_count, total_references, gaps[]

Shape

{ tools_tracked, gap_count, total_references,
  gaps: { slug, reference_count, referenced_by[] }[] }

Per-tool freshness: when each tool was last verified, bucketed into fresh / due / stale tiers aligned with the weekly scrape cadence. Mirrors /freshness. Sorted stalest-first.

Top-level keys

feed_version, generated_at, thresholds, buckets, tools[]

Shape

{ thresholds: { fresh_days, stale_days },
  buckets: { fresh, due, stale },
  tools: { slug, name, last_verified, age_days, staleness }[] }

Per-tool most-recent scrape attempt: outcome, plans found, warnings, error message on failure, plus cumulative attempt count. Pairs with /api/freshness.json (verified-data staleness) and /api/coverage.json (missing alternatives) so external monitors get three independent signals on data health.

Top-level keys

feed_version, generated_at, tools_with_runs, total_warnings_last_run, last_run_outcome_counts, runs[]

Shape

{ tools_with_runs, total_warnings_last_run,
  last_run_outcome_counts: { ok?, stale-preserved?, ... },
  runs: { slug, name, attempts,
    latest: null | { run_at, outcome, plans_found?, warnings?, error? } }[] }