What entered the system, and when
Retrieval and processing are separate. An explicit ingest writes what the feeds served to a dated capture file; everything else reads from those files. Nothing on this page polls, and nothing animates to suggest activity that is not happening.
- Retrieved
- 12
- Admittedpassed the 0.35 relevance threshold
- 12 (100%)
- Most recent item
- 02 Sep 17:30
- · telangana.gov.in/feed/ — parsed 10 items
- · telangana.gov.in/news/press-releases/feed/ — parsed 10 items
Government Orders, tabled figures or budget documents — the primary-document tier this system most needs
- Retrieved
- 124
- Admittedpassed the 0.35 relevance threshold
- 59 (48%)
- Most recent item
- 02 Sep 19:23
- × siasat.com/tag/revanth-reddy/feed/ — Chief Minister coverage: The operation was aborted due to timeout
- · siasat.com/feed/ — main feed: parsed 40 items
Full article text
- Retrieved
- 651
- Admittedpassed the 0.35 relevance threshold
- 45 (7%)
- Most recent item
- 02 Sep 19:24
- · deccanchronicle.com/rss_feed/southern-states — parsed 541 items
Telangana-specific curation — the rail is regional and mostly about elsewhere
- Retrieved
- 69
- Admittedpassed the 0.35 relevance threshold
- 60 (87%)
- Most recent item
- 02 Sep 18:35
- · thehansindia.com/rss/telangana — Telangana desk: parsed 60 items
Any Telugu-language reporting — this outlet publishes in English
- Retrieved
- 94
- Admittedpassed the 0.35 relevance threshold
- 21 (22%)
- Most recent item
- 02 Sep 19:26
- · ntvtelugu.com/category/telangana/feed/ — Telangana: parsed 60 items
- · ntvtelugu.com/feed/ — main feed: parsed 60 items
Broadcast transcripts, running order or airtime — this is the channel's website, not the channel
760 of 957 retrieved records fell below the relevance threshold and are held without contributing to any figure. Most are national or other-state items carried on a regional rail — a road accident in Assam, a flood in Nepal — which name no Telangana entity or place.
15 sources are not connected at all. Two name this system in their robots.txt and are excluded outright; one blocks AI crawlers without naming us and is honoured anyway; the rest need a licence or expose no traceable URL.
Capture files are written once and not modified. A conclusion drawn from a record last week can still be checked against the bytes that were retrieved, after the feed itself has moved on.