open/llm-ingestion-pipeline-security
1
0
Fork 0
Code Issues Releases Activity

docs(limitations): two variables moved between the 2400 and 2401 counts, not one

Browse source 
The previous commit said the two corpus counts differ by harvester over the same
knowledge base. Checked the provenance rather than leaving it asserted: 2401
traces to ms-ai-architect's 2026-07-26 message, 2400 to their 07-31 re-run, so
the attribution was right -- but the knowledge base was not held constant. Their
file count moved 389 -> 394 between the runs.

That inverts how the delta reads. A larger snapshot returning one fewer distinct
URL is not a rounding difference between two scripts; two variables moved at
once, so neither count explains the other and the one-URL gap is not evidence of
anything in particular.

Strengthens the do-not-merge rule instead of weakening it. The earlier wording
would have let a reader treat the counts as near-identical measurements of one
thing, which is the merge it was written to prevent.
This commit is contained in:
Kjell Tore Guttormsen 2026-07-31 15:43:34 +02:00
commit e25de560f9
1 changed files with 4 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

6
docs/LIMITATIONS.md
View file

@ -178,8 +178,10 @@ items; this is the full list, each with the mechanism.
(`is_ordinary_url` is False on the `%` rule) and still not a reason to move the
threshold — but the margin this bullet reports is much thinner than 4.04 implies.
Note the corpus count: 2400 is a rebuilt harvester's, where the 2401 in the next
bullet is the earlier run's. Same knowledge base, different scripts, one URL apart —
they are not interchangeable and must not be summed or quoted as one figure.
bullet is the earlier run's. The difference is not only the script — the corpus also
grew from 389 files to 394 between the two runs, so the *larger* snapshot returned one
*fewer* distinct URL. Two variables moved at once, which means neither count explains
the other: they are not interchangeable, and must not be summed or quoted as one figure.
- **Legitimate CDN content-asset ids trip the hex branch, permanently.** A ≥32-char
hex path segment is opaque by design, and several public CMSes mint asset URLs that
way (measured: 9 distinct on `regjeringen.no`, `ks.no`, `datatilsynet.no`). This is