chore: scrub private data from repo; honest doc tweaks

- Remove eval/corpus/{sessions,text-blocks}.json from the repo — they are real captured transcripts (file paths, project names, message content). gitignore eval/corpus/; local copies kept for the eval scripts.

- gen_sweep.mjs: replace a hardcoded /Users/<name> absolute import with a relative path.

- README: 'holds ~92k readable chars' -> 'up to ~92k chars' (don't overclaim legibility at the new page density).

- CHANGELOG: note OCR legibility at the larger page size is not independently re-eval'd.

.gitignore

@@ -42,3 +42,6 @@ SESSION_PLAYBOOK.md
 .continue/
 eval/gsm8k/novel_imgs/
 eval/gist-recall/work*/*.png
+# Eval corpus = real captured transcripts (PII: file paths, project names, and
+# message content). Keep a local copy for eval scripts, but never commit it.
+eval/corpus/

CHANGELOG.md

@@ -18,7 +18,10 @@ multi-agent code review with five confirmed fixes. Reviewed at extra-high recall
   1568→1932; dense tool/history pages now `DENSE_CONTENT_COLS=384` /
   `DENSE_CONTENT_CHARS_PER_IMAGE=92160` (1928×1928 full page) — fewer image
   blocks at the same OCR-validated 5×8 cell. The static slab is unchanged
-  (313 cols / 1573×1280). Pages never trip a server-side downscale.
+  (313 cols / 1573×1280). Pages never trip a server-side downscale. Note: the
+  larger per-page density uses the validated 5×8 cell and stays within
+  Anthropic's pixel/token limits, but OCR legibility at this page size has not
+  been independently re-eval'd (revert = the four render constants).
 - **Opus is OFF by default.** Production scope defaults to **Fable-5 only**;
   Opus 4.8/4.7 are opt-in (they read imaged content at a measurable tax — see
   FINDINGS.md). Opt in via `PXPIPE_MODELS` or the dashboard chips.

README.md

@@ -179,8 +179,8 @@ blocks, splices them back cache-friendly (static prefix preserved, so prompt
 caching keeps working), and forwards. Per-request events log to
 `~/.pxpipe/events.jsonl`.
 
-The economics: a 1928×1928 image costs ≈4,761 vision tokens and holds ≈92,000
-readable chars (≈48,000 text tokens at the observed density), so plain text is
+The economics: a 1928×1928 image costs ≈4,761 vision tokens and holds up to
+≈92,000 chars (≈48,000 text tokens at the observed density), so plain text is
 cheaper *only* when it runs denser than ~19 chars/token. Claude Code transcripts
 are far below that (observed 1.91 chars/token, N=391). The runtime estimator (`estimateImageCount`) plus a chars/token gate
 decides per-request; sparse prose is left as text.

eval/glyph-matrix/sweep/gen_sweep.mjs

@@ -2,7 +2,7 @@
 // Isolates ONE variable — pixels-per-glyph (cell size) — from density/search.
 // Short lines so every cell size fits one <=1568px page. Same content across
 // sizes => only resolution changes. Reader accuracy vs cell size = the curve.
-import { renderTextToPngs } from '/Users/user/Downloads/repos/pixelpipe/dist/core/render.js';
+import { renderTextToPngs } from '../../../dist/core/render.js';
 import { writeFileSync, mkdirSync } from 'node:fs';
 const OUT = '/tmp/sweep'; mkdirSync(OUT, { recursive: true });