@stupify/cli 0.0.16 → 0.1.0

package/src/model.ts

@@ -1,367 +0,0 @@
-import { execFile, spawn } from "node:child_process";
-import {
-  mkdir,
-  open,
-  readFile,
-  rename,
-  rm,
-  stat,
-  writeFile,
-} from "node:fs/promises";
-import { homedir, platform } from "node:os";
-import path from "node:path";
-import { promisify } from "node:util";
-import { MODEL_REGISTRY } from "./constants.ts";
-import type { ModelId } from "./types.ts";
-import type { CliUi } from "./ui.ts";
-
-const execFileAsync = promisify(execFile);
-const LLAMA_SERVER_HOST = "127.0.0.1";
-
-export type ModelProfile = "scout";
-
-type ModelRuntime = Readonly<{
-  profile: ModelProfile;
-  baseUrl: string;
-  port: string;
-  contextSize: number;
-  reasoning: "on" | "off" | "auto";
-  reasoningBudget?: number;
-  gpuLayers?: number;
-  batchSize?: number;
-  ubatchSize?: number;
-  parallel?: number;
-  threads?: number;
-  threadsBatch?: number;
-  flashAttention?: boolean;
-}>;
-
-export type LocalModel = Readonly<{
-  id: ModelId;
-  name: string;
-  baseUrl: string;
-  profile: ModelProfile;
-}>;
-
-export async function firstRunModelBootstrap(
-  modelId: ModelId,
-  ui: CliUi,
-): Promise<string> {
-  const selectedModel = MODEL_REGISTRY[modelId];
-  const modelDir = path.join(cacheDir(), "models");
-  const modelPath = path.join(modelDir, selectedModel.file);
-  if (await exists(modelPath)) return modelPath;
-
-  ui.note(`No local Stupify model found.
-Stupify runs locally.
-Download this model now?
-Model: ${selectedModel.name}
-Size: ${selectedModel.size}`, "Setup", { force: true });
-
-  if (!(await ui.confirm("Continue?"))) throw new Error("Setup cancelled.");
-
-  await mkdir(modelDir, { recursive: true });
-  await downloadModel(modelPath, selectedModel.url, ui);
-  if (!(await exists(modelPath)))
-    throw new Error("Model download failed: file was not created.");
-  return modelPath;
-}
-
-export async function loadLocalModel(
-  modelPath: string,
-  modelId: ModelId,
-  profile: ModelProfile,
-  ui: CliUi,
-): Promise<LocalModel> {
-  const selectedModel = MODEL_REGISTRY[modelId];
-  const runtime = modelRuntime(profile);
-  const runningModel = await runningServerModel(runtime.baseUrl);
-
-  if (runningModel) {
-    if (runningModel !== modelId) await stopManagedServer(runtime, ui);
-    if (runningModel === modelId) {
-      ui.info(`Using local model: ${selectedModel.name}`);
-      return {
-        id: modelId,
-        name: selectedModel.name,
-        baseUrl: runtime.baseUrl,
-        profile,
-      };
-    }
-  }
-
-  await ensureLlamaServerBinary();
-  await startLlamaServer(modelPath, modelId, selectedModel.name, runtime, ui);
-  const ready = ui.spinner(`Waiting for local ${profile} model`);
-  try {
-    await waitForServer(runtime.baseUrl, modelId);
-    ready.stop(`Local ${profile} model ready`);
-  } catch (error) {
-    ready.error(`Local ${profile} model failed to start`);
-    throw error;
-  }
-  return {
-    id: modelId,
-    name: selectedModel.name,
-    baseUrl: runtime.baseUrl,
-    profile,
-  };
-}
-
-function modelRuntime(profile: ModelProfile): ModelRuntime {
-  const baseUrl =
-    process.env.STUPIFY_SCOUT_LLAMA_SERVER_URL ??
-    process.env.STUPIFY_LLAMA_SERVER_URL ??
-    "http://127.0.0.1:8091";
-  return {
-    profile,
-    baseUrl,
-    port: new URL(baseUrl).port || "8091",
-    contextSize: envInteger("STUPIFY_LLAMA_CONTEXT") ?? 65_536,
-    reasoning: "off",
-    gpuLayers: envInteger("STUPIFY_LLAMA_GPU_LAYERS") ?? 999,
-    batchSize: envInteger("STUPIFY_LLAMA_BATCH") ?? 2_048,
-    ubatchSize: envInteger("STUPIFY_LLAMA_UBATCH") ?? 512,
-    parallel: envInteger("STUPIFY_LLAMA_PARALLEL") ?? 2,
-    threads: envInteger("STUPIFY_LLAMA_THREADS"),
-    threadsBatch: envInteger("STUPIFY_LLAMA_THREADS_BATCH"),
-    flashAttention: envBoolean("STUPIFY_LLAMA_FLASH_ATTN"),
-  };
-}
-
-async function runningServerModel(baseUrl: string): Promise<string | null> {
-  try {
-    const response = await fetch(`${baseUrl}/v1/models`, {
-      signal: AbortSignal.timeout(500),
-    });
-    if (!response.ok) return null;
-    const body = (await response.json()) as { data?: Array<{ id?: unknown }> };
-    const id = body.data?.[0]?.id;
-    return typeof id === "string" ? id : null;
-  } catch {
-    return null;
-  }
-}
-
-async function ensureLlamaServerBinary(): Promise<void> {
-  try {
-    await execFileAsync("llama-server", ["--version"], {
-      maxBuffer: 1024 * 1024,
-    });
-  } catch {
-    throw new Error(`Stupify needs llama-server for local inference.
-Install llama.cpp first:
-  brew install llama.cpp`);
-  }
-}
-
-async function startLlamaServer(
-  modelPath: string,
-  modelId: ModelId,
-  modelName: string,
-  runtime: ModelRuntime,
-  ui: CliUi,
-): Promise<void> {
-  const logDir = path.join(cacheDir(), "logs");
-  await mkdir(logDir, { recursive: true });
-  const logPath = path.join(logDir, "llama-server.log");
-  const out = await open(logPath, "a");
-  const err = await open(logPath, "a");
-
-  ui.step(`Starting local model server: ${modelName}`);
-  ui.info(`llama-server log: ${logPath}`);
-
-  const args = [
-    "-m",
-    modelPath,
-    "-a",
-    modelId,
-    "--host",
-    LLAMA_SERVER_HOST,
-    "--port",
-    runtime.port,
-    "-c",
-    String(runtime.contextSize),
-    "--reasoning",
-    runtime.reasoning,
-    "--no-warmup",
-  ];
-  if (runtime.gpuLayers !== undefined) args.push("-ngl", String(runtime.gpuLayers));
-  if (runtime.batchSize !== undefined) args.push("-b", String(runtime.batchSize));
-  if (runtime.ubatchSize !== undefined) args.push("-ub", String(runtime.ubatchSize));
-  if (runtime.parallel !== undefined) args.push("-np", String(runtime.parallel));
-  if (runtime.threads !== undefined) args.push("-t", String(runtime.threads));
-  if (runtime.threadsBatch !== undefined) args.push("-tb", String(runtime.threadsBatch));
-  if (runtime.flashAttention !== undefined) args.push("-fa", runtime.flashAttention ? "on" : "off");
-  if (runtime.reasoningBudget !== undefined) {
-    args.push("--reasoning-budget", String(runtime.reasoningBudget));
-  }
-
-  const child = spawn("llama-server", args, {
-    detached: true,
-    stdio: ["ignore", out.fd, err.fd],
-  });
-
-  child.unref();
-  if (child.pid) await writeFile(pidPath(runtime), String(child.pid));
-  await out.close();
-  await err.close();
-}
-
-async function stopManagedServer(runtime: ModelRuntime, ui: CliUi): Promise<void> {
-  const pid = await managedServerPid(runtime);
-  if (!pid) {
-    const runningModel = await runningServerModel(runtime.baseUrl);
-    throw new Error(`A llama-server is already running with ${runningModel ?? "another model"}.
-Stop it before switching models, or use STUPIFY_LLAMA_SERVER_URL for that server.`);
-  }
-
-  ui.step("Restarting local model server for selected model.");
-  try {
-    process.kill(pid, "SIGTERM");
-  } catch {
-    await rm(pidPath(runtime), { force: true });
-    return;
-  }
-
-  const deadline = Date.now() + 15_000;
-  while (Date.now() < deadline) {
-    if (!(await runningServerModel(runtime.baseUrl))) {
-      await rm(pidPath(runtime), { force: true });
-      return;
-    }
-    await sleep(250);
-  }
-
-  throw new Error("Timed out while stopping existing llama-server.");
-}
-
-async function managedServerPid(runtime: ModelRuntime): Promise<number | null> {
-  try {
-    const value = Number((await readFile(pidPath(runtime), "utf8")).trim());
-    return Number.isInteger(value) && value > 0 ? value : null;
-  } catch {
-    return null;
-  }
-}
-
-function pidPath(_runtime: ModelRuntime): string {
-  return path.join(cacheDir(), "llama-server.pid");
-}
-
-function envInteger(name: string, fallback?: number): number | undefined {
-  const raw = process.env[name];
-  if (raw === undefined || raw === "") return fallback;
-  const value = Number(raw);
-  return Number.isInteger(value) && value > 0 ? value : fallback;
-}
-
-function envBoolean(name: string): boolean | undefined {
-  const raw = process.env[name];
-  if (raw === undefined || raw === "") return undefined;
-  return /^(1|true|yes|on)$/i.test(raw);
-}
-
-async function waitForServer(baseUrl: string, modelId: ModelId): Promise<void> {
-  const deadline = Date.now() + 120_000;
-  while (Date.now() < deadline) {
-    const runningModel = await runningServerModel(baseUrl);
-    if (runningModel === modelId) return;
-    await sleep(500);
-  }
-  throw new Error(`llama-server did not become ready for ${modelId}.`);
-}
-
-function sleep(ms: number): Promise<void> {
-  return new Promise((resolve) => setTimeout(resolve, ms));
-}
-
-async function downloadModel(
-  modelPath: string,
-  modelUrl: string,
-  ui: CliUi,
-): Promise<void> {
-  const tempPath = `${modelPath}.download`;
-  await rm(tempPath, { force: true });
-
-  const downloadSpinner = ui.spinner("Downloading model");
-  let downloadProgress: ReturnType<CliUi["progress"]> | null = null;
-  try {
-    const response = await fetch(modelUrl);
-    if (!response.ok || !response.body)
-      throw new Error(`Model download failed: HTTP ${response.status}`);
-
-    const total = Number(response.headers.get("content-length") ?? 0);
-    let received = 0;
-    let lastPrint = 0;
-    let lastProgressBytes = 0;
-    const reader = response.body.getReader();
-    const file = await open(tempPath, "wx");
-    if (total > 0) {
-      downloadSpinner.clear();
-      downloadProgress = ui.progress("Downloading model", total);
-    }
-
-    try {
-      while (true) {
-        const { done, value } = await reader.read();
-        if (done) break;
-        received += value.byteLength;
-        await file.write(value);
-        const now = Date.now();
-        if (total > 0 && now - lastPrint > 500) {
-          lastPrint = now;
-          downloadProgress?.advance(
-            received - lastProgressBytes,
-            `${formatBytes(received)} / ${formatBytes(total)}`,
-          );
-          lastProgressBytes = received;
-        }
-      }
-    } finally {
-      await file.close();
-    }
-
-    if (downloadProgress && received > lastProgressBytes) {
-      downloadProgress.advance(
-        received - lastProgressBytes,
-        `${formatBytes(received)} / ${formatBytes(total)}`,
-      );
-    }
-    const activeProgress = downloadProgress ?? downloadSpinner;
-    activeProgress.stop(
-      total > 0
-        ? `Downloaded ${formatBytes(received)} / ${formatBytes(total)}`
-        : "Downloaded model",
-    );
-    await rename(tempPath, modelPath);
-  } catch (error) {
-    const activeProgress = downloadProgress ?? downloadSpinner;
-    activeProgress.error("Model download failed");
-    await rm(tempPath, { force: true });
-    throw error;
-  }
-}
-
-function cacheDir(): string {
-  if (process.env.STUPIFY_CACHE_DIR) return process.env.STUPIFY_CACHE_DIR;
-  if (process.env.XDG_CACHE_HOME)
-    return path.join(process.env.XDG_CACHE_HOME, "stupify");
-  if (platform() === "darwin")
-    return path.join(homedir(), "Library", "Caches", "stupify");
-  if (platform() === "win32" && process.env.LOCALAPPDATA)
-    return path.join(process.env.LOCALAPPDATA, "stupify", "Cache");
-  return path.join(homedir(), ".cache", "stupify");
-}
-
-async function exists(filePath: string): Promise<boolean> {
-  try {
-    return (await stat(filePath)).isFile();
-  } catch {
-    return false;
-  }
-}
-
-function formatBytes(bytes: number): string {
-  return `${(bytes / 1024 / 1024).toFixed(1)} MB`;
-}

package/src/prompts.ts

@@ -1,100 +0,0 @@
-import type { SemChangeSet, SemContext, SemContextPack, StupifyCheck } from "./types.ts";
-
-export function searchPrompt(input: Readonly<{
-  changeSet: SemChangeSet;
-  contexts: readonly SemContext[];
-  pack: SemContextPack;
-  patterns: readonly StupifyCheck[];
-  includeCounterReason: boolean;
-}>): string {
-  return `You are Stupify's local search model.
-Stupify checks whether AI-assisted coding may be replacing developer judgment.
-You will receive:
-1. Semantic changed entities selected by a fast local counter.
-2. Compressed local file context from Repomix.
-3. A list of search targets. Each target has exactly one assigned pattern.
-
-Your job:
-Evaluate each target only against its assigned pattern.
-False positives are expensive.
-Only emit a match if the assigned pattern clearly applies to that exact target.
-Do not perform general code review.
-Do not suggest improvements.
-Do not choose a pattern.
-Do not apply other patterns.
-Do not report issues for unlisted targets.
-Do not emit clean results.
-Omitted target = clean.
-Return JSON only:
-{
-  "matches": [
-    {
-      "targetId": "t001",
-      "reason": "one sentence",
-      "proof": "short pointer"
-    }
-  ]
-}
-
-Rules:
-- Use only targetIds from the input.
-- Emit at most 5 matches.
-- Prefer omission over a weak match.
-- Do not quote source code.
-- Do not write generic feedback.
-- Do not emit "no evidence" or "does not apply."
-- Proof must point to concrete changed product code that implements the pattern.
-- Proof must not be a file header or start with "diff --git".
-- Do not use pattern registry text, prompt text, docs, tests, or examples as proof.
-- Do not treat pattern or prompt wording as the code being evaluated.
-- Do not treat plain conditionals, guard clauses, skip paths, or error handling as indirection.
-- For unnecessary_complexity, identify the exact new named abstraction in proof.
-- If unnecessary_complexity proof would only be a file, hunk, or conditional block, omit it.
-- If nothing clearly matches, return { "matches": [] }.
-
-SOURCE:
-${input.changeSet.label}
-
-SEARCH TARGETS:
-${input.contexts.map((context) => formatSearchTarget(context, patternForContext(context, input.patterns), input.includeCounterReason)).join("\n\n") || "(none)"}
-
-REPOMIX CONTEXT (${input.pack.filePaths.length} files, ${input.pack.totalTokens} tokens):
-${input.pack.text || "(none)"}`;
-}
-
-function formatSearchPattern(check: StupifyCheck): string {
-  return `Pattern: ${check.id} (${check.name})
-Why this matters: ${check.why}
-Question: ${check.searchPrompt ?? check.question}
-Look for:
-${check.lookFor.map((signal) => `- ${signal}`).join("\n")}
-Ignore when:
-${check.ignoreWhen.map((signal) => `- ${signal}`).join("\n")}
-Match examples:
-${(check.searchExamples?.match ?? check.examples?.match ?? []).map((example) => `- ${example}`).join("\n")}
-Non-match examples:
-${(check.searchExamples?.nonMatch ?? check.examples?.noMatch ?? []).map((example) => `- ${example}`).join("\n")}`;
-}
-
-function formatSearchTarget(context: SemContext, pattern: StupifyCheck, includeCounterReason: boolean): string {
-  return `TARGET ${context.targetId}
-ASSIGNED ${formatSearchPattern(pattern)}
-SEM TARGET:
-ENTITY ${context.entityId}
-NAME ${context.entityName}
-KIND ${context.entityKind}
-CHANGE ${context.changeKind}
-FILE ${context.filePath ?? "(unknown)"}
-${includeCounterReason ? `COUNTER_REASON ${context.reason}` : ""}`.trim();
-}
-
-function patternForContext(context: SemContext, patterns: readonly StupifyCheck[]): StupifyCheck {
-  return patterns.find((pattern) => pattern.id === context.checkId) ?? {
-    id: context.checkId,
-    name: context.checkId,
-    question: `Does this target match ${context.checkId}?`,
-    why: "This pattern may indicate judgment-offload.",
-    lookFor: [],
-    ignoreWhen: [],
-  };
-}