teamchong 59a416cb09 feat(history): always-on history compression + fix gate-vs-renderer asymmetry
Rip the compressHistory opt-in and run Variant C history-image collapse on
every request that has a closed-prefix run. There is now ONE codepath: no
toggle, no env switch, no per-request opt-in. The reason the option was
defaulted-off was a -250% live measurement on 2026-05-19 (commit c3506e3 era);
that wasn't a "marginal upside" problem, it was a math bug. This commit
fixes the bug and turns the feature on.

Root cause of the -250% loss
----------------------------
`isCompressionProfitable()` has two paths: a row-aware path (string input,
matches `renderTextToPngs` image budgeting exactly) and a looser chars-only
path (number input, assumes dense lines). History.ts was calling it with
`text.length` (number) → loose chars-only estimate. History text is
newline-heavy (`--- role ---` headers, JSON arg blocks, [tool_use] /
[tool_result] labels), so the chars-only estimate under-counted images by
5-10× and let net-losers through the gate.

Secondary contributor: tool_use args were serialised with
`JSON.stringify(input, null, 2)` — pretty-printed, one short row per JSON
field. At cols=100 each field claimed a full row instead of wrapping at
~100 chars, inflating image count for tool-heavy histories.

Changes
-------
- src/core/history.ts
  - `ProfitableFn` signature: `(text: string, cols: number)` (was `textLen: number`).
    Documented the asymmetry so callers can't regress this.
  - `blocksToText` tool_use: compact `JSON.stringify(input)`, no indent.
  - Pass `text` (not `text.length`) into the profitability check.
- src/core/transform.ts
  - Delete `compressHistory`, `historyKeepTail`, `historyMinPrefix` from
    TransformOptions and DEFAULTS.
  - History collapse is now unconditional when `messages[]` is non-empty.
  - Wrap the gate in a closure that pins `numCols=1` + per-request `cpt`,
    so the gate's row-aware decision is byte-identical to the single-col
    renderer's actual image count.
- src/worker.ts: drop `COMPRESS_HISTORY`, `HISTORY_KEEP_TAIL`,
  `HISTORY_MIN_PREFIX` env vars from `Env` + transform options.
- src/node.ts: stale comment refresh — no toggles, period.
- tests/history.test.ts:
  - Rename describe "transformRequest + compressHistory" →
    "transformRequest history compression (always-on)".
  - Remove every `compressHistory: true, historyKeepTail: X, historyMinPrefix: Y`
    from transformRequest calls — options no longer exist.
  - Bump fixtures for default `keepTail=4` / `minCollapsePrefix=10`
    (14-turn body, 3500 chars/turn) so the now-honest row-aware gate
    accepts. Renamed "8-closed + 2-live" → "10-closed + 4-live" to match
    the new defaults.
  - Bump straddle-test per-turn body 2500→3500 chars: tool block labels
    add ~65 chars of header overhead that pushed the tighter fixture
    under the row-aware boundary.
  - Rename "JSON-pretty args" test → "COMPACT JSON args"; assert no
    pretty indentation.

Verification
------------
- npm test → 263/263 green
- npm run typecheck → clean
- npm run build → dist/node.js builds clean
2026-05-19 23:54:11 -04:00
2026-05-18 11:52:02 -04:00

pixelpipe

A token-saving proxy for Claude Code that renders the system prompt + tool definitions as images, so Claude OCRs them instead of paying for them as text. 65-73% input-token savings on Opus 4.7, 100% reasoning quality preserved, identical fixed text every turn for a clean prompt-cache.

Runs on Node 18+ and Cloudflare Workers from the same source.


How it works

                                  ┌─ original ────────────────────┐
                                  │ ~68K input tok                │
Claude Code  ──►  pixelpipe  ──►  │  (system + tools as text)     │  ──►  Anthropic
                       │          └───────────────────────────────┘
                       └──────►   ┌─ via proxy ───────────────────┐
                                  │ ~3.5K input tok               │
                                  │  (system + tools as PNG +     │
                                  │   prompt-cache breakpoint)    │
                                  └───────────────────────────────┘
                                          ↓ Anthropic vision OCR
                                          100% reasoning quality retained

The proxy intercepts POST /v1/messages, pulls the system prompt + tool documentation out of the JSON body, renders it into one or more grayscale PNGs using a build-time-generated GNU Unifont glyph atlas (covers ~35k BMP codepoints by default — Latin, Cyrillic, Greek, CJK, Hiragana, Katakana, Hangul, Hebrew, Arabic, math symbols, box drawing, decorative symbols), and substitutes those PNGs back in as image content blocks with an ephemeral cache_control breakpoint.

Token math (Opus 4.7, real Claude Code workflow):

metric original via proxy savings
Cold input tokens ~68K ~3.5K 95%
Cache-warm input tokens ~7.5K ~3.5K 53%
Per-call median (mixed) - - 65-73%
Per-image OCR quality vs txt - - ~99.5%

Quick start (Node)

npm install
npm run build           # produces dist/node.js
node bin/cli.js         # listens on 127.0.0.1:47821 by default

After editing code, restart in one step:

pnpm run restart                              # graceful SIGTERM → rebuild → start
pnpm run restart -- --no-build                # skip rebuild (dist/ is fresh)
PORT=47822 pnpm run restart                   # override listen port via env

pnpm run restart does, in order:

  1. Lists every running pixelpipe PID (via pgrep) and SIGTERMs them all. Orphans from prior crashed sessions are cleaned up too.
  2. Waits up to 5s for graceful exit (the SIGTERM handler flushes the JSONL tracker). Escalates to SIGKILL only if anything's still alive.
  3. Runs pnpm run build. Build failures abort the restart — the script refuses to start a stale binary. Pass --no-build to skip when you know dist/ is fresh.
  4. Checks the target port is free. If it isn't, names the holding process and refuses to start (cheaper than a crashed Node stacktrace).
  5. execs node bin/cli.js in the foreground so Ctrl-C reaches Node. The proxy takes no behavioral flags — env vars only (see Configuration).

Point Claude Code at it:

ANTHROPIC_BASE_URL=http://127.0.0.1:47821 \
  claude --exclude-dynamic-system-prompt-sections

That's it. Use Claude Code normally.

The --exclude-dynamic-system-prompt-sections flag suppresses the small per-turn variable section so the rendered image stays byte-identical across turns — that's what makes the prompt cache actually hit.


Quick start (Cloudflare Workers)

npx wrangler dev        # local dev on :8787
npx wrangler deploy     # ship to *.workers.dev

Then in Claude Code:

ANTHROPIC_BASE_URL=https://pixelpipe.<your-account>.workers.dev \
  claude --exclude-dynamic-system-prompt-sections

You can attach a custom hostname and route in wrangler.toml.


Configuration

The proxy runs with a single codepath. Every compression mode is on, every break-even threshold is at its measured-best value, and tuning parameters are not user-adjustable. The only configurable surface is where to listen, what to proxy, and where to log — env-var only, no CLI flags.

env var default meaning
PORT 47821 Node only — listen port
ANTHROPIC_UPSTREAM https://api.anthropic.com upstream API base
PIXELPIPE_LOG ~/.pixelpipe/events.jsonl persistent event log

In Workers, set the optional upstream API key with:

npx wrangler secret put ANTHROPIC_API_KEY

If unset, the proxy forwards whatever x-api-key the client sent.


Architecture

src/
├── core/              100% runtime-agnostic (Web Standard APIs only)
│   ├── atlas.ts         (generated) sparse Unicode atlas, base64-inlined
│   ├── png.ts           minimal grayscale PNG encoder
│   ├── render.ts        text → PNG bytes
│   ├── transform.ts     request body rewriter
│   ├── proxy.ts         the fetch handler
│   └── types.ts         Anthropic API types
├── node.ts            node:http adapter + CLI
└── worker.ts          export default { fetch }

scripts/
├── gen-atlas.ts       build-time: OTF → atlas.ts (uses @napi-rs/canvas)
└── build.mjs          esbuild bundler for Node target

assets/
├── Unifont-16.0.04.otf       primary font (~35k BMP codepoints w/ full-bmp profile)
├── UNIFONT_LICENSE.txt       OFL + GPL-with-font-exception
└── JetBrainsMono-Regular.ttf legacy / ASCII-only fallback (kept on disk)

The atlas is generated at build time from Unifont-16.0.04.otf, base64-inlined into a .ts file with sparse codepoint + offset tables (binary-packed), and shipped with the bundle. At runtime there are zero external files to read and zero non-Web-Standard imports — that's the only way this works in Workers without per-request asset fetches.

Regenerate the atlas (after swapping fonts, sizes, or codepoint profile):

pnpm run build:atlas                          # default: full-bmp (~35k cp, all BMP Unifont covers)
ATLAS_PROFILE=practical pnpm run build:atlas  # drops Hangul (~24k cp; for Workers free-tier)
FONT_PX=12 pnpm run build:atlas               # nondefault size; verify cells

Limitations

  • The bundled GNU Unifont at 10px (cell 5×11 px Latin, 10×11 CJK) is Anthropic-OCR-clean for ~35k BMP codepoints by default (full-bmp profile): Latin, Cyrillic, Greek, CJK Unified Ideographs, Hiragana, Katakana, Hangul, Hebrew, Arabic, math symbols, box-drawing, arrows, Dingbats, Letterlike Symbols, Enclosed Alphanumerics, etc. Drops for codepoints outside the profile (e.g. emoji 😀 — supplementary plane) get counted in events.jsonl#dropped_chars (with the top-20 broken out as dropped_codepoints_top) so you can spot patterns. For Workers free-tier deployments under the 1 MB compressed-bundle cap, switch to ATLAS_PROFILE=practical pnpm run build:atlas (~24k cp; drops Hangul). Right-to-left scripts render left-to-right in source order (no bidi shaping); Devanagari / Thai / similar complex shaping is also unsupported.
  • Compression sets a 5-minute prompt-cache TTL. Adding cache_control: ephemeral causes warm-cache rotation, not eviction.
  • A 5KB break-even point: if input is < MIN_COMPRESS_CHARS chars we skip compression entirely (overhead would exceed savings).
  • Per-machine font: regenerate the atlas if you swap fonts. The generated src/core/atlas.ts is checked in so consumers don't need @napi-rs/canvas to install.
  • Workers CPU limit: this is fine for free-tier (10ms CPU) on small prompts; large prompts (>30K chars) may need the paid tier.

Development

npm install
npm run dev:node              # tsx watch on src/node.ts
npm run dev:worker            # wrangler dev
npm run test                  # vitest
npm run test:watch
npm run typecheck             # tsc --noEmit
pnpm run build:atlas          # regenerate src/core/atlas.ts from OTF
npm run build                 # build dist/node.js
npm run deploy:worker         # wrangler deploy

License

MIT.

S
Description
cut Fable 5 token usage by rendering text context as images
Readme MIT 24 MiB
Languages
TypeScript 86.7%
JavaScript 9.5%
Shell 2.1%
Python 1.7%