Independent fixed public-seed hash audit — 17 October 2026

PASS. No arithmetic, source-mapping, corpus, or reported-count error found.
The exact scripts were rerun with Python's isolated mode and standard library
in a private mirror. Both regenerated JSON objects equal the originals.
The outcome remains zero matches within the explicitly tested finite family.

Corpus and controls
- All 102 entries are distinct 20-byte seed values. Their union equals the
  65 distinct saved Certicom challenge seeds, 27 saved standard seed records,
  and all 23 SEC2 v1 S values, with overlaps deduplicated.
- All 23 SEC2 seeds were matched to their named curve definitions. Each of
  the ten seeded DER files matches its same named primary SEC2 section.
  This strengthens the producer's global-set membership check.
- The published ECC2-131 seedP hashes to the reported SHA-1 digest and its
  low 131 bits equal the published U_x exactly. The target starts and field
  polynomials match the portable fixture. Target integers have 125 and
  182 bits respectively; 131 and 193 are the field degrees.

Counts independently recomputed
- 24,794 raw seed-neighborhood inputs plus 408 alternate input/mode pairs.
- 100,808 primary hash evaluations; another 24,794 successor SHA-1 calls
  support expansion, plus one historical control. Import self-tests extra.
- 2,771,404 coordinate recipe comparisons, including duplicate values.
- 5,443,632 exact 152-bit interior comparisons, excluded from that count.

Interpretation limits
The negative result applies only to these public seeds, raw offsets -2..255,
four alternate encodings at offset zero, four digest byte/word layouts, and
the exact reductions/expansion/interior rules in the script. Retained literal
name metadata is not searched by this script. Single 128/160-bit hashes cannot
equal the 182-bit target; only the expansion/remainder tests reach its size.
Interior-window matches would be partial candidates, not complete recipes.
Unknown seeds/messages, different algorithms, output increments and other
conversions remain untested. Fuzzy visual overlap supports no conclusion.

Sources: saved primary text, not a fresh download or independent OCR.
https://www.secg.org/SEC2-Ver-1.0.pdf
https://www.certicom.com/content/certicom/en/curves-list.html
https://safecurves.cr.yp.to/grouper.ieee.org/groups/1363/private/x9-62-09-20-98.pdf

Receipt: fixed-public-hash-audit.json (source mapping, counts, file hashes).
Reproducer: audit_fixed_hash_check.py (stdlib only; writes only this folder).
