Skip to content

armgit / ECC

Public Imported from GitHub
0

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

13.4 KB
Raw Blame
'use strict';
/**
 * Tests for scripts/lib/transcript-context.js (#2155)
 *
 * Covers transcript usage extraction, context-window detection, threshold and
 * interval resolution, and the bucket math the strategic-compact hook uses.
 *
 * Run with: node tests/lib/transcript-context.test.js
 */

const assert = require('assert');
const fs = require('fs');
const os = require('os');
const path = require('path');
const {
  STANDARD_CONTEXT_WINDOW_TOKENS,
  LARGE_CONTEXT_WINDOW_TOKENS,
  DEFAULT_CONTEXT_THRESHOLD_STANDARD,
  DEFAULT_CONTEXT_THRESHOLD_LARGE,
  DEFAULT_CONTEXT_INTERVAL_TOKENS,
  DEFAULT_TRANSCRIPT_TAIL_BYTES,
  readFileTail,
  readLatestContextTokens,
  resolveContextWindowTokens,
  resolveContextThreshold,
  resolveContextInterval,
  computeContextBucket,
  formatWindowLabel,
  isContextWindowInferred
} = require('../../scripts/lib/transcript-context');

console.log('=== Testing transcript-context.js ===\n');

let passed = 0;
let failed = 0;

function test(desc, fn) {
  try {
    fn();
    console.log(`  ✓ ${desc}`);
    passed++;
  } catch (e) {
    console.log(`  ✗ ${desc}: ${e.message}`);
    failed++;
  }
}

const transcriptRoot = fs.mkdtempSync(path.join(os.tmpdir(), 'ecc-transcript-context-'));
process.on('exit', () => fs.rmSync(transcriptRoot, { recursive: true, force: true }));
let fixtureSeq = 0;

function writeTranscript(lines) {
  fixtureSeq += 1;
  const filePath = path.join(transcriptRoot, `transcript-context-test-${process.pid}-${fixtureSeq}.jsonl`);
  fs.writeFileSync(filePath, lines.join('\n') + '\n', { flag: 'wx', mode: 0o600 });
  return filePath;
}

function usageRecord(tokens, model = 'claude-sonnet-4-6', extra = {}) {
  return JSON.stringify({
    type: 'assistant',
    message: {
      model,
      usage: {
        input_tokens: tokens.input || 0,
        cache_read_input_tokens: tokens.cacheRead || 0,
        cache_creation_input_tokens: tokens.cacheCreation || 0,
        output_tokens: tokens.output || 0
      }
    },
    ...extra
  });
}

const cleanupPaths = [];

function tracked(filePath) {
  cleanupPaths.push(filePath);
  return filePath;
}

// ── readLatestContextTokens ──
console.log('readLatestContextTokens:');

test('sums input + cache_read + cache_creation from the latest usage record', () => {
  const file = tracked(writeTranscript([usageRecord({ input: 10, cacheRead: 20, cacheCreation: 5 }), usageRecord({ input: 100, cacheRead: 150000, cacheCreation: 7000 })]));
  const result = readLatestContextTokens(file);
  assert.ok(result, 'Expected a usage result');
  assert.strictEqual(result.tokens, 157100);
});

test('returns the model id alongside the token count', () => {
  const file = tracked(writeTranscript([usageRecord({ input: 1000 }, 'claude-opus-4-5[1m]')]));
  const result = readLatestContextTokens(file);
  assert.strictEqual(result.model, 'claude-opus-4-5[1m]');
});

test('skips trailing records without usage (e.g. tool results)', () => {
  const file = tracked(writeTranscript([usageRecord({ input: 5000 }), JSON.stringify({ type: 'user', message: { content: 'tool result' } }), JSON.stringify({ type: 'system', subtype: 'info' })]));
  const result = readLatestContextTokens(file);
  assert.strictEqual(result.tokens, 5000);
});

test('skips malformed JSONL lines without throwing', () => {
  const file = tracked(writeTranscript([usageRecord({ input: 4200 }), '{not json at all', '']));
  const result = readLatestContextTokens(file);
  assert.strictEqual(result.tokens, 4200);
});

test('returns null for a transcript with no usage records', () => {
  const file = tracked(writeTranscript([JSON.stringify({ type: 'user', message: { content: 'hello' } })]));
  assert.strictEqual(readLatestContextTokens(file), null);
});

test('returns null for a missing transcript file', () => {
  assert.strictEqual(readLatestContextTokens(path.join(transcriptRoot, 'definitely-missing.jsonl')), null);
});

test('returns null for empty or non-string paths', () => {
  assert.strictEqual(readLatestContextTokens(''), null);
  assert.strictEqual(readLatestContextTokens(undefined), null);
});

test('ignores zero-token usage records', () => {
  const file = tracked(writeTranscript([usageRecord({ input: 999 }), usageRecord({ input: 0 })]));
  const result = readLatestContextTokens(file);
  assert.strictEqual(result.tokens, 999);
});

test('only scans the transcript tail (latest records win on large files)', () => {
  const filler = JSON.stringify({ type: 'system', note: 'x'.repeat(512) });
  const lines = [usageRecord({ input: 11 })];
  for (let i = 0; i < 50; i++) lines.push(filler);
  lines.push(usageRecord({ input: 170000 }));
  const file = tracked(writeTranscript(lines));
  // Tail window smaller than the file forces the truncated-tail path.
  const result = readLatestContextTokens(file, { tailBytes: 4096 });
  assert.strictEqual(result.tokens, 170000);
});

// ── resolveContextWindowTokens ──
console.log('\nresolveContextWindowTokens:');

// Isolation: an env-set window override (either knob) otherwise leaks into the
// default-window assertions below and fails them (#2290).
const originalContextWindowEnv = {
  ECC_CONTEXT_WINDOW_TOKENS: process.env.ECC_CONTEXT_WINDOW_TOKENS,
  CLAUDE_CODE_AUTO_COMPACT_WINDOW: process.env.CLAUDE_CODE_AUTO_COMPACT_WINDOW,
};
delete process.env.ECC_CONTEXT_WINDOW_TOKENS;
delete process.env.CLAUDE_CODE_AUTO_COMPACT_WINDOW;

test('defaults to the standard 200k window', () => {
  assert.strictEqual(resolveContextWindowTokens(50000, 'claude-sonnet-4-6'), STANDARD_CONTEXT_WINDOW_TOKENS);
});

test('honors an explicit ECC_CONTEXT_WINDOW_TOKENS override (e.g. 400k models, #2290)', () => {
  process.env.ECC_CONTEXT_WINDOW_TOKENS = '400000';
  try {
    assert.strictEqual(resolveContextWindowTokens(50000, 'claude-opus-4-x'), 400000);
  } finally {
    delete process.env.ECC_CONTEXT_WINDOW_TOKENS;
  }
});

test('honors Claude Code native CLAUDE_CODE_AUTO_COMPACT_WINDOW override', () => {
  process.env.CLAUDE_CODE_AUTO_COMPACT_WINDOW = '400000';
  try {
    assert.strictEqual(resolveContextWindowTokens(50000, 'claude-opus-4-x'), 400000);
  } finally {
    delete process.env.CLAUDE_CODE_AUTO_COMPACT_WINDOW;
  }
});

test('ignores a non-positive / invalid window override', () => {
  process.env.ECC_CONTEXT_WINDOW_TOKENS = 'not-a-number';
  try {
    assert.strictEqual(resolveContextWindowTokens(50000, 'claude-sonnet-4-6'), STANDARD_CONTEXT_WINDOW_TOKENS);
  } finally {
    delete process.env.ECC_CONTEXT_WINDOW_TOKENS;
  }
});

test('detects a 1M window from the [1m] model marker', () => {
  assert.strictEqual(resolveContextWindowTokens(50000, 'claude-opus-4-5[1m]'), LARGE_CONTEXT_WINDOW_TOKENS);
});

test('detects a 1M window when observed tokens exceed 200k (marker dropped)', () => {
  assert.strictEqual(resolveContextWindowTokens(220000, 'claude-opus-4-5'), LARGE_CONTEXT_WINDOW_TOKENS);
});

test('recognizes claude-fable-5 as a 1M window without a [1m] marker or 200k+ tokens (#2461)', () => {
  assert.strictEqual(resolveContextWindowTokens(187000, 'claude-fable-5'), LARGE_CONTEXT_WINDOW_TOKENS);
});

test('recognizes claude-mythos-5 as a 1M window from the known-model table (#2461)', () => {
  assert.strictEqual(resolveContextWindowTokens(50000, 'claude-mythos-5'), LARGE_CONTEXT_WINDOW_TOKENS);
});

test('recognizes claude-opus-5 as a 1M window from the known-model table', () => {
  assert.strictEqual(resolveContextWindowTokens(50000, 'claude-opus-5'), LARGE_CONTEXT_WINDOW_TOKENS);
});

test('recognizes dated/prefixed variants of known large-window model ids (#2461)', () => {
  assert.strictEqual(resolveContextWindowTokens(50000, 'us.anthropic.claude-fable-5-20260115-v1:0'), LARGE_CONTEXT_WINDOW_TOKENS);
});

test('env window override still wins over the known-model table (#2461)', () => {
  process.env.ECC_CONTEXT_WINDOW_TOKENS = '400000';
  try {
    assert.strictEqual(resolveContextWindowTokens(50000, 'claude-fable-5'), 400000);
  } finally {
    delete process.env.ECC_CONTEXT_WINDOW_TOKENS;
  }
});

test('does not match hypothetical smaller tiers sharing a known-family prefix (#2461)', () => {
  assert.strictEqual(resolveContextWindowTokens(50000, 'claude-fable-5-mini'), STANDARD_CONTEXT_WINDOW_TOKENS);
  assert.strictEqual(resolveContextWindowTokens(50000, 'claude-mythos-5-haiku-20260201'), STANDARD_CONTEXT_WINDOW_TOKENS);
});

test('keeps the 200k default for unknown model ids at low token counts (no false positives, #2461)', () => {
  assert.strictEqual(resolveContextWindowTokens(187000, 'claude-haiku-4-5-20251001'), STANDARD_CONTEXT_WINDOW_TOKENS);
});

test('treats an empty model id as standard window', () => {
  assert.strictEqual(resolveContextWindowTokens(100000, ''), STANDARD_CONTEXT_WINDOW_TOKENS);
});

// ── isContextWindowInferred ──
console.log('\nisContextWindowInferred:');

test('flags the assumed 200k default as inferred', () => {
  assert.strictEqual(isContextWindowInferred(187000, 'claude-opus-9'), true);
});

test('an env override is a detected window, not inferred', () => {
  process.env.ECC_CONTEXT_WINDOW_TOKENS = '1000000';
  try {
    assert.strictEqual(isContextWindowInferred(187000, 'claude-opus-9'), false);
  } finally {
    delete process.env.ECC_CONTEXT_WINDOW_TOKENS;
  }
});

test('a [1m] marker is a detected window, not inferred', () => {
  assert.strictEqual(isContextWindowInferred(187000, 'claude-opus-4-5[1m]'), false);
});

test('a known large-window family is a detected window, not inferred', () => {
  assert.strictEqual(isContextWindowInferred(187000, 'claude-fable-5'), false);
});

test('tokens above the standard window still leave the exact size inferred', () => {
  assert.strictEqual(isContextWindowInferred(220000, 'claude-opus-9'), true);
});

for (const [name, value] of Object.entries(originalContextWindowEnv)) {
  if (value === undefined) delete process.env[name];
  else process.env[name] = value;
}

// ── resolveContextThreshold ──
console.log('\nresolveContextThreshold:');

test('defaults to 160k for the 200k window', () => {
  assert.strictEqual(resolveContextThreshold({}, STANDARD_CONTEXT_WINDOW_TOKENS), DEFAULT_CONTEXT_THRESHOLD_STANDARD);
});

test('defaults to 250k for the 1M window', () => {
  assert.strictEqual(resolveContextThreshold({}, LARGE_CONTEXT_WINDOW_TOKENS), DEFAULT_CONTEXT_THRESHOLD_LARGE);
});

test('honours COMPACT_CONTEXT_THRESHOLD override', () => {
  assert.strictEqual(resolveContextThreshold({ COMPACT_CONTEXT_THRESHOLD: '1234' }, STANDARD_CONTEXT_WINDOW_TOKENS), 1234);
});

test('COMPACT_CONTEXT_THRESHOLD=0 disables the signal', () => {
  assert.strictEqual(resolveContextThreshold({ COMPACT_CONTEXT_THRESHOLD: '0' }, STANDARD_CONTEXT_WINDOW_TOKENS), 0);
});

test('invalid COMPACT_CONTEXT_THRESHOLD falls back to the default', () => {
  for (const bad of ['-5', 'abc', '99999999999']) {
    assert.strictEqual(resolveContextThreshold({ COMPACT_CONTEXT_THRESHOLD: bad }, STANDARD_CONTEXT_WINDOW_TOKENS), DEFAULT_CONTEXT_THRESHOLD_STANDARD, `Expected fallback for ${bad}`);
  }
});

// ── resolveContextInterval ──
console.log('\nresolveContextInterval:');

test('defaults to 60k tokens', () => {
  assert.strictEqual(resolveContextInterval({}), DEFAULT_CONTEXT_INTERVAL_TOKENS);
});

test('honours COMPACT_CONTEXT_INTERVAL override', () => {
  assert.strictEqual(resolveContextInterval({ COMPACT_CONTEXT_INTERVAL: '5000' }), 5000);
});

test('invalid COMPACT_CONTEXT_INTERVAL falls back to the default', () => {
  for (const bad of ['0', '-1', 'abc']) {
    assert.strictEqual(resolveContextInterval({ COMPACT_CONTEXT_INTERVAL: bad }), DEFAULT_CONTEXT_INTERVAL_TOKENS, `Expected fallback for ${bad}`);
  }
});

// ── computeContextBucket ──
console.log('\ncomputeContextBucket:');

test('returns -1 below the threshold', () => {
  assert.strictEqual(computeContextBucket(159999, 160000, 60000), -1);
});

test('returns bucket 0 at the threshold', () => {
  assert.strictEqual(computeContextBucket(160000, 160000, 60000), 0);
});

test('increments the bucket after each interval of growth', () => {
  assert.strictEqual(computeContextBucket(219999, 160000, 60000), 0);
  assert.strictEqual(computeContextBucket(220000, 160000, 60000), 1);
  assert.strictEqual(computeContextBucket(280000, 160000, 60000), 2);
});

test('returns -1 when the threshold is disabled (0)', () => {
  assert.strictEqual(computeContextBucket(500000, 0, 60000), -1);
});

test('returns -1 for non-finite token counts', () => {
  assert.strictEqual(computeContextBucket(NaN, 160000, 60000), -1);
});

// ── readFileTail ──
console.log('\nreadFileTail:');

test('is exported and returns the trailing bytes with a truncation flag', () => {
  assert.strictEqual(typeof readFileTail, 'function');
  assert.strictEqual(DEFAULT_TRANSCRIPT_TAIL_BYTES, 256 * 1024);
  const file = tracked(writeTranscript(['first-line', 'second', 'third']));
  const whole = readFileTail(file, 1024);
  assert.deepStrictEqual(whole, { text: 'first-line\nsecond\nthird\n', truncated: false });
  const tail = readFileTail(file, 6);
  assert.deepStrictEqual(tail, { text: 'third\n', truncated: true });
  assert.strictEqual(readFileTail(path.join(transcriptRoot, `missing-${process.pid}.jsonl`), 10), null);
});

// ── formatWindowLabel ──
console.log('\nformatWindowLabel:');

test('labels the standard and large windows', () => {
  assert.strictEqual(formatWindowLabel(STANDARD_CONTEXT_WINDOW_TOKENS), '200k');
  assert.strictEqual(formatWindowLabel(LARGE_CONTEXT_WINDOW_TOKENS), '1M');
});

// Cleanup
for (const filePath of cleanupPaths) {
  try {
    fs.unlinkSync(filePath);
  } catch {
    /* ignore */
  }
}

console.log(`\nResults: Passed: ${passed}, Failed: ${failed}`);
process.exit(failed > 0 ? 1 : 0);