SIGNAL & CONTEXT: OFFLINE RSS / SOURCE-LEDGER EXPERIMENT Executed 2026-11-06. This is a small parser check, not a crawler service or a fact-checking system. FILES signal-rss-ledger-check-20261106.py.txt Original URDREAM example Python script signal-rss-ledger-requirements-20261106.txt Exact tested package versions signal-rss-ledger-public-feed-20261106.xml.txt Unmodified public RSS snapshot signal-rss-ledger-fixtures-20261106.json Ten explicitly synthetic inputs signal-rss-ledger-results-20261106.json Recorded outputs and 22 checks signal-rss-ledger-fetch-20261106.json Public HTTP capture receipt, allowlisted headers only REPLAY Place the seven files in one directory. The tested interpreter was Python 3.13.7. The dependency requires Python >=3.10; other interpreters were not tested. Create a disposable environment, install the pinned packages, then execute the text script: python3 -m venv .venv .venv/bin/python -m pip install -r signal-rss-ledger-requirements-20261106.txt .venv/bin/python signal-rss-ledger-check-20261106.py.txt --public-feed signal-rss-ledger-public-feed-20261106.xml.txt --fixtures signal-rss-ledger-fixtures-20261106.json --output replay-results.json --expected-results signal-rss-ledger-results-20261106.json Installing dependencies needs package-network access. The experiment itself reads only saved files. Expected terminal summary: PASS, 22/22 checks, 18 public entries, replay_core_matches_expected=true. The 'outcomes' object should match exactly. New timestamps and local timings will differ. The script intentionally asserts this fixed snapshot's 18 items; a future live feed can change. INPUT / IDENTITY BOUNDARY Public input: https://signal.urdream.top/rss.xml, captured 2026-11-06T10:38:47.027112+00:00. 9,144 bytes, SHA256 6c62435cdf6edf107031b103d16acf894cfa9b43b05c3d232fd173f1004ab6d8. Two anonymous GETs occurred; the initial identical capture lacked a completed environment receipt because an obsolete dependency name was queried. The second capture has the full safe receipt. No cookies, authentication, private account data, visitor data or linked articles were requested. The raw snapshot contains only public URDREAM feed metadata/summaries and has not been redacted or modified. Synthetic fixtures use reserved .example hosts. No synthetic URL is fetched. The repeated-public-input case parses the same saved bytes twice; it is not a second live observation. The malformed XML deliberately omits its closing tags. Recovery of an entry does not make it trustworthy. WHAT THE EXAMPLE ADDS feedparser parses feeds. Our short editorial_ledger function separately demonstrates grouping by feed URL + item ID (or exact link), retaining every observation, detecting changes in selected parsed fields and holding malformed/non-feed batches. Those are our example decisions, not built-in product deduplication or factual review. No URL query normalization is performed. Missing identities remain separate observations. A common link does not establish independent corroboration. HASH / TIMING BOUNDARY run.public_feed_sha256 hashes the captured XML, not any linked article. parsed_fields_sha256 hashes the selected normalized entry fields, not raw source-document bytes. Linked-document hashes and resolved document URLs are null because retrieval was NOT_RUN. Server ETag remains a separate opaque value. The results include 25 local cached-byte parse timings on Darwin/arm64, Python 3.13.7. These exclude network latency, are not a VPS/load benchmark and do not establish a production resource budget. NOT RUN Linked-document retrieval, article fact verification, semantic-change acceptance, actual editorial review, conditional GET/304, unattended scheduling, production ingestion/publication, AI/models, VPS load, browser behavior or network-proxy impact. No library source, environment or cache is bundled. DEPENDENCIES / SOURCES feedparser 6.0.14: BSD-2-Clause; https://github.com/kurtmckee/feedparser License: https://github.com/kurtmckee/feedparser/blob/develop/LICENSE Documentation: https://feedparser.readthedocs.io/en/latest/ feedparser-sgmllib 2.1.0: PSF-2.0 per installed package metadata; https://github.com/python-syndication/feedparser-sgmllib The two packages are obtained independently from their publishers; none of their code is redistributed here. Original example code/fixtures and the public URDREAM snapshot are supplied as inspectable research artifacts. No output establishes Google indexing, AdSense approval, model quality or real-provider behavior.