Files
OmniRoute/open-sse/services/compression/lite.ts
Diego Rodrigues de Sa e Souza 7db430a352 Release v3.8.14 (#3340)
* chore(release): open v3.8.14 development cycle

Version bump 3.8.13 -> 3.8.14 (root + electron + open-sse + openapi + lockfiles).
Seed the v3.8.14 changelog with the four post-tag hotfixes that shipped to
Docker/Electron in v3.8.13 but missed the immutable npm 3.8.13 (#3336 SSRF /
CodeQL #323, #3334/#3335/#3339 Electron packaging). i18n CHANGELOG mirrors get
the in-progress placeholder section.

* feat: add per-provider custom headers support for OpenAI/Anthropic-compatible nodes (#3338)

Integrated into release/v3.8.14

* fix: Kiro Builder ID token import fails with Bad credentials (#3333)

Integrated into release/v3.8.14 — adds Builder ID cached-creds + OIDC refresh path for Kiro token import, with regression tests (#3333).

* Improve code quality: auto-pr/docstrings-1780792063 (#3337)

Integrated into release/v3.8.14 — docstring for context analytics route re-export.

* fix(catalog): remove minimaxai/minimax-m3 from NVIDIA NIM tier (404 upstream) (#3329) (#3341)

NVIDIA NIM does not host minimaxai/minimax-m3 — every request returns
404 page not found, while sibling minimaxai/minimax-m2.7 on the same provider
works. Advertising a model that 404s is a catalog bug; remove it from the nvidia
tier (it remains on the tiers that actually serve MiniMax M3). Re-add only once
NVIDIA serves it.

Co-authored-by: mikmaneggahommie <mikmaneggahommie@users.noreply.github.com>

* fix(cli): write OpenCode config to ~/.config on all platforms incl. Windows (#3330) (#3343)

resolveOpencodeConfigDir used %APPDATA% on Windows, but OpenCode reads its
config from XDG ~/.config/opencode/ on every platform (on Windows:
%USERPROFILE%\.config\opencode\, NOT %APPDATA%). So a Windows user who
configured OpenCode via the dashboard had the file written where OpenCode never
looks — it silently had no effect.

Use the XDG path (XDG_CONFIG_HOME || ~/.config) unconditionally. Update the UI
note + route JSDoc, and flip the three tests that encoded the old %APPDATA%
behavior (t40 per-platform + card-note, cli-runtime-extended getCliConfigPaths).

Co-authored-by: abdulkadirozyurt <abdulkadirozyurt@users.noreply.github.com>

* fix(proxy): make auto-selection fallback opt-in (#3332) (#3344)

selectWorkingProxyFallback (Step 11 of resolveProxyForConnection) listed ALL
registry proxies, ignoring assignments and per-connection proxy_enabled, and
returned the first working one with level:'autoSelect'. So a single proxy added
to the registry silently became a global fallback for every connection's traffic.

Gate it behind a new PROXY_AUTO_SELECT_ENABLED feature flag (default off): the
fallback now no-ops unless the operator opts in. No registry proxy becomes a
silent global default anymore.

Co-authored-by: hertznsk <hertznsk@users.noreply.github.com>

* fix(sse): treat MiniMax M3 as multimodal so vision isn't stripped (#3328) (#3342)

MiniMax M3 via the opencode provider (oc/minimax-m3-free) appeared blind:
image inputs didn't reach the model, while the same model in Cline could
see them. Verified empirically that MiniMax M3 on the opencode upstream IS
multimodal -- a base64 image is described correctly (it returns 403 only
for remote image URLs, which it doesn't accept).

Root cause: OmniRoute treated MiniMax M3 as a non-vision model in two
places, so when compression was active the image was replaced with a text
placeholder before dispatch:
- compression's modelSupportsVision() heuristic (lite.ts) only matched
  gpt-4/4o/claude-3/gemini/vision -- minimax was absent -> replaceImageUrls
  stripped the image.
- the opencode minimax-m3-free catalog entry lacked supportsVision, so the
  combo vision-capability gate could also exclude/mishandle it.

Add 'minimax-m3' to the vision heuristic and supportsVision: true to the
opencode minimax-m3-free entry. TDD: a failing-then-passing test in
compression/lite.test.ts proves replaceImageUrls now keeps images for
minimax-m3 ids, plus a registry assertion mirroring the #2822 qwen test.

Reported-by: @mikmaneggahommie

* docs(i18n): translate 25 core documentation files to Indonesian (#3348)

Integrated into release/v3.8.14 — Indonesian i18n docs.

* fix(review): resolve /review-reviews battery findings (LEDGER-1..11) on v3.8.14 (#3350)

Integrated into release/v3.8.14 — /review-reviews battery hardening (LEDGER-1..11) for #3338 custom-headers + #3333 kiro, plus cycle-test drift fixes (#3329/#3330/#3332).

* fix(provider-proxy): honor per-account proxy toggles (#3349)

Integrated into release/v3.8.14 — honor per-account proxy toggles + auto-fallback opt-in via PROXY_AUTO_SELECT_ENABLED.

* fix(dashboard): remove duplicate Distribute Proxies button on provider page (#3352)

* fix(providers): reduce proxy label noise (#3346)

Integrated into release/v3.8.14 — reduce proxy label noise + a11y (aria-label/sr-only).

* fix(duckduckgo): restore bare Response contract and rebase onto release/v3.8.14 (#3323)

Integrated into release/v3.8.14 — browser-backed cookie providers (duckduckgo/claude-web) with restored executor contract + unit tests.

* fix(noauth): expose only usable model aliases (#3345)

Integrated into release/v3.8.14 — noauth usable-alias filtering + registry alias plumbing (veo-free).

* fix(dashboard): stop infinite config-load loop on Hermes Agent detail page (#3353)

* fix(electron): tree-kill the server on exit/update to release the omniroute.exe lock (#3347) (#3354)

* chore(release): finalize v3.8.14 changelog + clear release-gate drift

- CHANGELOG: finalize the v3.8.14 section (date, full New Features/Bug Fixes/
  Maintenance coverage of all 16 cycle commits, Contributors hall of 12).
- docs: document OMNIROUTE_BROWSER_POOL + WEB_COOKIE_USE_BROWSER (#3323) in
  .env.example + ENVIRONMENT.md; regenerate the id/llm.txt strict mirror (#3348
  had translated it; llm.txt mirrors must match root).
- test(proxy-fetch): #3323 made tlsClient.available a computed getter — stub it
  via Object.defineProperty instead of assignment (5 tests were red on the base).

* fix(translator): coerce Gemini functionDeclaration parameters to an OBJECT schema (#3357) (#3360)

* fix(gemini): resolve truncation/suppression of false positive textual tool call markers in backticks (#3358)

Integrated into release/v3.8.14 — Gemini/Antigravity textual tool-call marker normalization (no false-positive suppression + split-chunk buffering).

* docs(changelog): add #3358 Gemini textual tool-call normalization to v3.8.14

* fix(dashboard): surface real analytics error instead of generic placeholder (#3356) (#3361)

The Analytics page discarded the server's error body on a non-OK response and
rendered a generic "An error occurred", so users (and maintainers) could not see
why /api/usage/analytics 500'd after an upgrade. Now the route returns the real
reason via buildErrorBody (sanitized, Hard Rule #12) and the page surfaces it via
a new readFetchErrorMessage helper that handles both the OpenAI-style and legacy
error shapes.

Reported-by: @superti4r

---------

Co-authored-by: PizzaV <103120356+pizzav-xyz@users.noreply.github.com>
Co-authored-by: Someres <168349709+quanturbo@users.noreply.github.com>
Co-authored-by: Dong Mengzhe <154944819+Lang-Qiu@users.noreply.github.com>
Co-authored-by: mikmaneggahommie <mikmaneggahommie@users.noreply.github.com>
Co-authored-by: abdulkadirozyurt <abdulkadirozyurt@users.noreply.github.com>
Co-authored-by: hertznsk <hertznsk@users.noreply.github.com>
Co-authored-by: Krisna Santosa <54174372+KrisnaSantosa15@users.noreply.github.com>
Co-authored-by: Randi <55005611+rdself@users.noreply.github.com>
Co-authored-by: Wilson <pedbookmed@gmail.com>
Co-authored-by: Paijo <14921983+oyi77@users.noreply.github.com>
Co-authored-by: Ardem2025 <ardemb22@gmail.com>
2026-06-07 07:20:02 -03:00

246 lines
7.0 KiB
TypeScript

import type { CompressionResult, CompressionMode } from "./types.ts";
import { createCompressionStats } from "./stats.ts";
interface Message {
role: string;
content: string | Array<Record<string, unknown>>;
[key: string]: unknown;
}
interface ChatBody {
messages?: Message[];
[key: string]: unknown;
}
interface LiteCompressionOptions {
model?: string;
supportsVision?: boolean | null;
preserveSystemPrompt?: boolean;
}
function trimTrailingHorizontalWhitespace(line: string): string {
let end = line.length;
while (end > 0) {
const code = line.charCodeAt(end - 1);
if (code !== 32 && code !== 9) break;
end--;
}
return end === line.length ? line : line.slice(0, end);
}
function collapseNewlineRuns(content: string): string {
let normalized = "";
let newlineRun = 0;
for (const char of content) {
if (char === "\n") {
newlineRun++;
if (newlineRun <= 2) {
normalized += char;
}
continue;
}
newlineRun = 0;
normalized += char;
}
return normalized;
}
function normalizeMessageWhitespace(content: string): string {
return collapseNewlineRuns(content).split("\n").map(trimTrailingHorizontalWhitespace).join("\n");
}
function modelSupportsVision(model: string): boolean {
const normalized = model.toLowerCase();
return (
normalized.includes("vision") ||
normalized.includes("gpt-4") ||
normalized.includes("4o") ||
normalized.includes("claude-3") ||
normalized.includes("gemini") ||
// #3328: MiniMax M3 is multimodal — verified it describes images via the opencode
// upstream. Without this, compression strips the image and the model goes "blind".
normalized.includes("minimax-m3")
);
}
export function collapseWhitespace(
body: ChatBody,
options: LiteCompressionOptions = {}
): {
body: ChatBody;
applied: boolean;
} {
if (!body.messages) return { body, applied: false };
let applied = false;
const messages = body.messages.map((msg) => {
if (options.preserveSystemPrompt === true && msg.role === "system") return msg;
if (typeof msg.content !== "string") return msg;
const normalized = normalizeMessageWhitespace(msg.content);
if (normalized !== msg.content) applied = true;
return { ...msg, content: normalized };
});
return { body: { ...body, messages }, applied };
}
export function dedupSystemPrompt(
body: ChatBody,
options: LiteCompressionOptions = {}
): {
body: ChatBody;
applied: boolean;
} {
if (!body.messages) return { body, applied: false };
if (options.preserveSystemPrompt === true) return { body, applied: false };
const seen = new Set<string>();
let applied = false;
const messages = body.messages.filter((msg) => {
if (msg.role !== "system" || typeof msg.content !== "string") return true;
const key = msg.content.trim().slice(0, 200);
if (seen.has(key)) {
applied = true;
return false;
}
seen.add(key);
return true;
});
return { body: { ...body, messages }, applied };
}
export function compressToolResults(body: ChatBody): {
body: ChatBody;
applied: boolean;
} {
if (!body.messages) return { body, applied: false };
const MAX_TOOL_LENGTH = 2000;
let applied = false;
const messages = body.messages.map((msg) => {
if (msg.role !== "tool" || typeof msg.content !== "string") return msg;
if (msg.content.length <= MAX_TOOL_LENGTH) return msg;
applied = true;
return {
...msg,
content: msg.content.slice(0, MAX_TOOL_LENGTH) + "\n...[truncated]",
};
});
return { body: { ...body, messages }, applied };
}
export function removeRedundantContent(
body: ChatBody,
options: LiteCompressionOptions = {}
): {
body: ChatBody;
applied: boolean;
} {
if (!body.messages) return { body, applied: false };
let applied = false;
const messages: Message[] = [];
for (let i = 0; i < body.messages.length; i++) {
const msg = body.messages[i];
if (options.preserveSystemPrompt === true && msg.role === "system") {
messages.push(msg);
continue;
}
const contentStr = typeof msg.content === "string" ? msg.content : JSON.stringify(msg.content);
if (
i > 0 &&
body.messages[i - 1].role === msg.role &&
typeof body.messages[i - 1].content === "string" &&
body.messages[i - 1].content === contentStr
) {
applied = true;
continue;
}
messages.push(msg);
}
return { body: { ...body, messages }, applied };
}
export function replaceImageUrls(
body: ChatBody,
options?: LiteCompressionOptions | string
): { body: ChatBody; applied: boolean } {
if (!body.messages) return { body, applied: false };
const supportsVision =
typeof options === "object" && options !== null
? options.supportsVision
: typeof options === "string"
? modelSupportsVision(options)
: undefined;
if (supportsVision !== false) return { body, applied: false };
let applied = false;
const messages = body.messages.map((msg) => {
if (!Array.isArray(msg.content)) return msg;
const newContent = msg.content.map((part) => {
if (
typeof part === "object" &&
part !== null &&
part.type === "image_url" &&
typeof (part as Record<string, unknown>).image_url === "object" &&
((part as Record<string, unknown>).image_url as Record<string, unknown>)?.url
) {
const url = String(
((part as Record<string, unknown>).image_url as Record<string, unknown>).url
);
if (url.startsWith("data:image/")) {
applied = true;
const format = url.slice(url.indexOf("/") + 1, url.indexOf(";")) || "unknown";
return { type: "text", text: `[image: ${format}]` };
}
}
return part;
});
return { ...msg, content: newContent };
});
return { body: { ...body, messages }, applied };
}
export function applyLiteCompression(
body: Record<string, unknown>,
options?: LiteCompressionOptions
): CompressionResult {
const originalBody = body;
let current = body as ChatBody;
const techniquesApplied: string[] = [];
const r1 = collapseWhitespace(current, options);
current = r1.body;
if (r1.applied) techniquesApplied.push("whitespace");
const r2 = dedupSystemPrompt(current, options);
current = r2.body;
if (r2.applied) techniquesApplied.push("system-dedup");
const r3 = compressToolResults(current);
current = r3.body;
if (r3.applied) techniquesApplied.push("tool-compress");
const r4 = removeRedundantContent(current, options);
current = r4.body;
if (r4.applied) techniquesApplied.push("redundant-remove");
const r5 = replaceImageUrls(current, options);
current = r5.body;
if (r5.applied) techniquesApplied.push("image-placeholder");
const compressed = techniquesApplied.length > 0;
const stats = compressed
? createCompressionStats(
originalBody,
current as Record<string, unknown>,
"lite" as CompressionMode,
techniquesApplied
)
: null;
return {
body: current as Record<string, unknown>,
compressed,
stats,
};
}