Sources: opencollective-accounts reads the 100 most active (200 took longer than the 20 s fetch limit) (#66)

ddiff --git a/README.md b/README.md
iindex 6ae24ea..9bd609a 100644
---- a/README.md
++++ b/README.md
@@@ -152,7 +152,7 @@ catalog is about 12 MB.
  
  `request_content_type` (`application/json` only) is sent as the Content-Type of the endpoint's GET, which has no body.
  Open Collective's public GraphQL API answers a GET only with it (its CSRF guard): `opencollective-accounts` (category
--`crowdfunding`) reads the 200 most active and the 100 newest collectives, projects and funds every six hours, each a
++`crowdfunding`) reads the 100 most active and the 100 newest collectives, projects and funds every six hours, each a
  `campaign` item with the amounts and counts as Open Collective states them, for OpenVibe.Deals' crowdfunding discovery.
  
  **Facts stated in prose (`extract`, rss/atom only).** A feed endpoint may carry
ddiff --git a/seeds/sources.json b/seeds/sources.json
iindex 080f0fd..d422cdd 100644
---- a/seeds/sources.json
++++ b/seeds/sources.json
@@@ -1246,7 +1246,7 @@
        "homepage_url": "https://opencollective.com/discover",
        "endpoints": [
          {
--          "url": "https://api.opencollective.com/graphql/v2?query=%7B+accounts%28type%3A+%5BCOLLECTIVE%2C+PROJECT%2C+FUND%5D%2C+limit%3A+200%2C+orderBy%3A+%7Bfield%3A+ACTIVITY%2C+direction%3A+DESC%7D%29+%7B+nodes+%7B+slug+name+type+description+currency+createdAt+imageUrl+website+tags+stats+%7B+totalAmountReceived+%7B+valueInCents+currency+%7D+yearlyBudget+%7B+valueInCents+currency+%7D+contributorsCount+%7D+%7D+%7D+%7D",
++          "url": "https://api.opencollective.com/graphql/v2?query=%7B+accounts%28type%3A+%5BCOLLECTIVE%2C+PROJECT%2C+FUND%5D%2C+limit%3A+100%2C+orderBy%3A+%7Bfield%3A+ACTIVITY%2C+direction%3A+DESC%7D%29+%7B+nodes+%7B+slug+name+type+description+currency+createdAt+imageUrl+website+tags+stats+%7B+totalAmountReceived+%7B+valueInCents+currency+%7D+yearlyBudget+%7B+valueInCents+currency+%7D+contributorsCount+%7D+%7D+%7D+%7D",
            "format": "json",
            "request_content_type": "application/json",
            "items_path": "data.accounts.nodes",
@@@ -1297,12 +1297,12 @@
          }
        ],
        "robots_note": "robots.txt checked 2026-10-15: https://api.opencollective.com/robots.txt answers 404, so the API host publishes no robots rules (a 404 is treated as no restrictions, never guessed as a disallow). The fetcher re-checks robots.txt on every run.",
--      "terms_note": "Open Collective's public GraphQL API v2 (documented at https://graphql-docs-v2.opencollective.com) answers anonymous reads of public accounts; a GET must carry Content-Type: application/json (its CSRF guard). Two requests every six hours: the 200 most active collectives, projects and funds and the 100 newest. Each becomes a campaign item: name, description, type, currency, the total received and yearly budget in cents, the number of contributors, logo URL, website and first tag, as Open Collective states them; the profile is https://opencollective.com/<slug>. Show amounts as observations with their fetch date, link each to its Open Collective page, and imply no endorsement.",
++      "terms_note": "Open Collective's public GraphQL API v2 (documented at https://graphql-docs-v2.opencollective.com) answers anonymous reads of public accounts; a GET must carry Content-Type: application/json (its CSRF guard). Two requests every six hours: the 100 most active collectives, projects and funds and the 100 newest. Each becomes a campaign item: name, description, type, currency, the total received and yearly budget in cents, the number of contributors, logo URL, website and first tag, as Open Collective states them; the profile is https://opencollective.com/<slug>. Show amounts as observations with their fetch date, link each to its Open Collective page, and imply no endorsement.",
        "license_note": "Public account data published by Open Collective; the descriptions are the collectives' own words (short summaries only, linked back).",
        "min_interval_ms": 60000,
        "poll_interval_sec": 21600,
        "stale_after_sec": 86400,
--      "max_items": 300,
++      "max_items": 200,
        "enabled": false,
        "review_required": true,
        "sensitivity": "financial",