diff --git a/scripts/hooks/cost-tracker.js b/scripts/hooks/cost-tracker.js index 3de1eaaec..981f67619 100755 --- a/scripts/hooks/cost-tracker.js +++ b/scripts/hooks/cost-tracker.js @@ -70,19 +70,37 @@ function readHarnessCost(sessionId, maxAgeSeconds) { // Approximate per-1M-token billing rates (USD). // Cache creation: 1.25x input rate. Cache read: 0.1x input rate. +// Source: https://platform.claude.com/docs/en/about-claude/pricing +// Current-generation list prices: Fable/Mythos 5 $10/$50, Opus 5 and +// Opus 4.5-4.8 $5/$25, Sonnet 5 $2/$10, Sonnet 4.6 $3/$15, and Haiku 4.5 +// $1/$5. Opus 4.0/4.1 and Opus 3 stay on the legacy $15/$75 tier. const RATE_TABLE = { - haiku: { in: 0.80, out: 4.0, cacheWrite: 1.00, cacheRead: 0.08 }, - sonnet: { in: 3.00, out: 15.0, cacheWrite: 3.75, cacheRead: 0.30 }, - opus: { in: 15.00, out: 75.0, cacheWrite: 18.75, cacheRead: 1.50 } + haiku: { in: 1.00, out: 5.0, cacheWrite: 1.25, cacheRead: 0.10 }, + sonnet: { in: 3.00, out: 15.0, cacheWrite: 3.75, cacheRead: 0.30 }, + sonnet5: { in: 2.00, out: 10.0, cacheWrite: 2.50, cacheRead: 0.20 }, + opus: { in: 5.00, out: 25.0, cacheWrite: 6.25, cacheRead: 0.50 }, + opusLegacy: { in: 15.00, out: 75.0, cacheWrite: 18.75, cacheRead: 1.50 }, + fable: { in: 10.00, out: 50.0, cacheWrite: 12.50, cacheRead: 1.00 } }; +// Opus 4.0's dated snapshot omits the minor segment, so an `opus-4-0` +// substring check alone misses `claude-opus-4-20250514`. +const LEGACY_OPUS_RE = /3-opus|opus-4-0(?!\d)|opus-4-1(?!\d)|opus-4[-@]\d{8}/; + function getRates(model) { const m = String(model || '').toLowerCase(); + if (m.includes('fable') || m.includes('mythos')) return RATE_TABLE.fable; if (m.includes('haiku')) return RATE_TABLE.haiku; + if (isSonnet5(m)) return RATE_TABLE.sonnet5; + if (LEGACY_OPUS_RE.test(m)) return RATE_TABLE.opusLegacy; if (m.includes('opus')) return RATE_TABLE.opus; return RATE_TABLE.sonnet; } +function isSonnet5(model) { + return /(?:^|[^a-z0-9])sonnet-5(?:[^a-z0-9]|$)/.test(model); +} + function toNumber(v) { const n = Number(v); return Number.isFinite(n) ? n : 0; diff --git a/scripts/hooks/session-end.js b/scripts/hooks/session-end.js index 7139562ad..9709aa95a 100644 --- a/scripts/hooks/session-end.js +++ b/scripts/hooks/session-end.js @@ -43,9 +43,16 @@ function extractSessionSummary(transcriptPath) { if (entry.type === 'user' || entry.role === 'user' || entry.message?.role === 'user') { // Support both direct content and nested message.content (Claude Code JSONL format) const rawContent = entry.message?.content ?? entry.content; + // Skip tool_result carrier turns — they are not user asks. + const isToolResult = Array.isArray(rawContent) && rawContent.some(c => c && c.type === 'tool_result'); const text = typeof rawContent === 'string' ? rawContent : Array.isArray(rawContent) ? rawContent.map(c => (c && c.text) || '').join(' ') : ''; const cleaned = stripAnsi(text).trim(); - if (cleaned) { + // Skip harness noise: local command echoes, caveats, system reminders. + const isNoise = /^<(local-command-caveat|local-command-stdout|command-name|command-message|command-args|system-reminder|task-notification)/i.test(cleaned); + // `isMeta` is also used for genuine channel- and plugin-originated + // human prompts. Exclude known structured noise above instead of + // discarding every metadata-marked user turn. + if (cleaned && !isToolResult && !isNoise) { userMessages.push(cleaned.slice(0, 200)); } } diff --git a/scripts/lib/cost-estimate.js b/scripts/lib/cost-estimate.js deleted file mode 100644 index a1651a8c9..000000000 --- a/scripts/lib/cost-estimate.js +++ /dev/null @@ -1,32 +0,0 @@ -'use strict'; - -/** - * Shared cost estimation for ECC hooks. - * - * Approximate per-1M-token blended rates (conservative defaults). - */ - -const RATE_TABLE = { - haiku: { in: 0.8, out: 4.0 }, - sonnet: { in: 3.0, out: 15.0 }, - opus: { in: 15.0, out: 75.0 } -}; - -/** - * Estimate USD cost from token counts. - * @param {string} model - Model name (may contain "haiku", "sonnet", or "opus") - * @param {number} inputTokens - * @param {number} outputTokens - * @returns {number} Estimated cost in USD (rounded to 6 decimal places) - */ -function estimateCost(model, inputTokens, outputTokens) { - const normalized = String(model || '').toLowerCase(); - let rates = RATE_TABLE.sonnet; - if (normalized.includes('haiku')) rates = RATE_TABLE.haiku; - if (normalized.includes('opus')) rates = RATE_TABLE.opus; - - const cost = (inputTokens / 1_000_000) * rates.in + (outputTokens / 1_000_000) * rates.out; - return Math.round(cost * 1e6) / 1e6; -} - -module.exports = { estimateCost, RATE_TABLE }; diff --git a/tests/hooks/cost-tracker.test.js b/tests/hooks/cost-tracker.test.js index 8f652d7a1..78e72f68c 100644 --- a/tests/hooks/cost-tracker.test.js +++ b/tests/hooks/cost-tracker.test.js @@ -54,6 +54,49 @@ function runScript(input, envOverrides = {}) { return { code: result.status || 0, stdout: result.stdout || '', stderr: result.stderr || '' }; } +function removeHarnessCostCache(sessionId) { + const cachePath = path.join(os.tmpdir(), `harness-cost-${sessionId}.json`); + try { + fs.unlinkSync(cachePath); + } catch (err) { + if (err.code !== 'ENOENT') throw err; + } +} + +function assertSonnet5CacheCost(cacheUsage, expectedCost, description) { + const tmpHome = makeTempDir(); + const sessionId = `sonnet5-${description}-${process.pid}-${Date.now()}`; + const transcriptPath = path.join(tmpHome, 'session.jsonl'); + writeTranscript(transcriptPath, [{ + type: 'assistant', + message: { + id: `msg_sonnet5_${description}`, + model: 'claude-sonnet-5', + usage: { + input_tokens: 1_000_000, + output_tokens: 1_000_000, + ...cacheUsage, + }, + }, + }]); + + try { + removeHarnessCostCache(sessionId); + const result = runScript( + { session_id: sessionId, transcript_path: transcriptPath }, + withTempHome(tmpHome) + ); + assert.strictEqual(result.code, 0, `Expected exit code 0, got ${result.code}`); + + const metricsFile = path.join(tmpHome, '.claude', 'metrics', 'costs.jsonl'); + const row = JSON.parse(fs.readFileSync(metricsFile, 'utf8').trim()); + assert.strictEqual(row.estimated_cost_usd, expectedCost, description); + } finally { + removeHarnessCostCache(sessionId); + fs.rmSync(tmpHome, { recursive: true, force: true }); + } +} + function runTests() { console.log('\n=== Testing cost-tracker.js ===\n'); @@ -297,7 +340,245 @@ function runTests() { } }) ? passed++ : failed++); - // 9. Ignores stale harness-cost cache and falls back to transcript estimate + // 9. Prices Sonnet 5 at the documented $2/$10 rate. + (test('prices Sonnet 5 at $12 per 1M input + 1M output tokens', () => { + const tmpHome = makeTempDir(); + const sessionId = `sonnet5-${process.pid}-${Date.now()}`; + const transcriptPath = path.join(tmpHome, 'session.jsonl'); + writeTranscript(transcriptPath, [ + { + type: 'assistant', + message: { + id: 'msg_sonnet5', + model: 'claude-sonnet-5', + usage: { input_tokens: 1_000_000, output_tokens: 1_000_000 }, + }, + }, + ]); + + fs.writeFileSync( + path.join(os.tmpdir(), `harness-cost-${sessionId}.json`), + JSON.stringify({ ts: Math.floor(Date.now() / 1000), cost_usd: 999 }), + 'utf8' + ); + + try { + removeHarnessCostCache(sessionId); + const result = runScript( + { session_id: sessionId, transcript_path: transcriptPath }, + withTempHome(tmpHome) + ); + assert.strictEqual(result.code, 0, `Expected exit code 0, got ${result.code}`); + + const metricsFile = path.join(tmpHome, '.claude', 'metrics', 'costs.jsonl'); + const row = JSON.parse(fs.readFileSync(metricsFile, 'utf8').trim()); + assert.strictEqual(row.estimated_cost_usd, 12, 'Expected Sonnet 5 1M/1M to cost $12.00'); + } finally { + removeHarnessCostCache(sessionId); + fs.rmSync(tmpHome, { recursive: true, force: true }); + } + }) ? passed++ : failed++); + + // 9b. Sonnet 5 cache write/read tokens use the correct rates. + (test('prices Sonnet 5 cache tokens at the documented rates', () => { + const tmpHome = makeTempDir(); + const sessionId = `sonnet5-cache-${process.pid}-${Date.now()}`; + const transcriptPath = path.join(tmpHome, 'session.jsonl'); + writeTranscript(transcriptPath, [ + { + type: 'assistant', + message: { + id: 'msg_sonnet5_cache', + model: 'claude-sonnet-5', + usage: { + input_tokens: 1_000_000, + output_tokens: 1_000_000, + cache_creation_input_tokens: 1_000_000, + cache_read_input_tokens: 1_000_000, + }, + }, + }, + ]); + + try { + removeHarnessCostCache(sessionId); + const result = runScript( + { session_id: sessionId, transcript_path: transcriptPath }, + withTempHome(tmpHome) + ); + assert.strictEqual(result.code, 0, `Expected exit code 0, got ${result.code}`); + + const metricsFile = path.join(tmpHome, '.claude', 'metrics', 'costs.jsonl'); + const row = JSON.parse(fs.readFileSync(metricsFile, 'utf8').trim()); + assert.strictEqual(row.estimated_cost_usd, 14.7, 'Expected Sonnet 5 1M input + 1M output + 1M cache write + 1M cache read to cost $14.70'); + } finally { + removeHarnessCostCache(sessionId); + fs.rmSync(tmpHome, { recursive: true, force: true }); + } + }) ? passed++ : failed++); + + // 9c. Cache write/read rates are independently covered. + (test('prices Sonnet 5 cache writes at $2.50 per 1M tokens', () => { + assertSonnet5CacheCost( + { cache_creation_input_tokens: 1_000_000 }, + 14.5, + 'cache-write' + ); + }) ? passed++ : failed++); + + (test('prices Sonnet 5 cache reads at $0.20 per 1M tokens', () => { + assertSonnet5CacheCost( + { cache_read_input_tokens: 1_000_000 }, + 12.2, + 'cache-read' + ); + }) ? passed++ : failed++); + + // 10. Sonnet 4.6 keeps the existing $3/$15 rate and is not mistaken for Sonnet 5. + (test('prices Sonnet 4.6 at $18 per 1M input + 1M output tokens', () => { + const tmpHome = makeTempDir(); + const sessionId = `sonnet46-${process.pid}-${Date.now()}`; + const transcriptPath = path.join(tmpHome, 'session.jsonl'); + writeTranscript(transcriptPath, [ + { + type: 'assistant', + message: { + id: 'msg_sonnet46', + model: 'claude-sonnet-4-6', + usage: { input_tokens: 1_000_000, output_tokens: 1_000_000 }, + }, + }, + ]); + + try { + removeHarnessCostCache(sessionId); + const result = runScript( + { session_id: sessionId, transcript_path: transcriptPath }, + withTempHome(tmpHome) + ); + assert.strictEqual(result.code, 0, `Expected exit code 0, got ${result.code}`); + + const metricsFile = path.join(tmpHome, '.claude', 'metrics', 'costs.jsonl'); + const row = JSON.parse(fs.readFileSync(metricsFile, 'utf8').trim()); + assert.strictEqual(row.estimated_cost_usd, 18, 'Expected Sonnet 4.6 1M/1M to remain $18.00'); + } finally { + removeHarnessCostCache(sessionId); + fs.rmSync(tmpHome, { recursive: true, force: true }); + } + }) ? passed++ : failed++); + + // 10b. Dated Sonnet 5 IDs are matched correctly. + (test('prices dated Sonnet 5 IDs at $12', () => { + const tmpHome = makeTempDir(); + const sessionId = `sonnet5-dated-${process.pid}-${Date.now()}`; + const transcriptPath = path.join(tmpHome, 'session.jsonl'); + writeTranscript(transcriptPath, [ + { + type: 'assistant', + message: { + id: 'msg_sonnet5_dated', + model: 'claude-sonnet-5-20261001', + usage: { input_tokens: 1_000_000, output_tokens: 1_000_000 }, + }, + }, + ]); + + try { + removeHarnessCostCache(sessionId); + const result = runScript( + { session_id: sessionId, transcript_path: transcriptPath }, + withTempHome(tmpHome) + ); + assert.strictEqual(result.code, 0, `Expected exit code 0, got ${result.code}`); + + const metricsFile = path.join(tmpHome, '.claude', 'metrics', 'costs.jsonl'); + const row = JSON.parse(fs.readFileSync(metricsFile, 'utf8').trim()); + assert.strictEqual(row.estimated_cost_usd, 12, 'Expected dated Sonnet 5 1M/1M to cost $12.00'); + } finally { + removeHarnessCostCache(sessionId); + fs.rmSync(tmpHome, { recursive: true, force: true }); + } + }) ? passed++ : failed++); + + // 10c. Near-miss Sonnet 5 IDs fall back to standard Sonnet rates. + (test('rejects claude-sonnet-50 as a Sonnet 5 near-miss', () => { + const tmpHome = makeTempDir(); + const sessionId = `sonnet50-near-miss-${process.pid}-${Date.now()}`; + const transcriptPath = path.join(tmpHome, 'session.jsonl'); + writeTranscript(transcriptPath, [ + { + type: 'assistant', + message: { + id: 'msg_sonnet50', + model: 'claude-sonnet-50', + usage: { input_tokens: 1_000_000, output_tokens: 1_000_000 }, + }, + }, + ]); + + try { + removeHarnessCostCache(sessionId); + const result = runScript( + { session_id: sessionId, transcript_path: transcriptPath }, + withTempHome(tmpHome) + ); + assert.strictEqual(result.code, 0, `Expected exit code 0, got ${result.code}`); + + const metricsFile = path.join(tmpHome, '.claude', 'metrics', 'costs.jsonl'); + const row = JSON.parse(fs.readFileSync(metricsFile, 'utf8').trim()); + assert.strictEqual(row.estimated_cost_usd, 18, 'Expected claude-sonnet-50 near-miss to fall back to $18.00 Sonnet rate'); + } finally { + removeHarnessCostCache(sessionId); + fs.rmSync(tmpHome, { recursive: true, force: true }); + } + }) ? passed++ : failed++); + + // 10d. Opus 4.0's dated ID has no explicit minor segment. It must retain + // the legacy $15/$75 rate while Opus 4.5 uses the current $5/$25 rate. + (test('distinguishes the dated Opus 4.0 snapshot from current Opus 4.x', () => { + const priceModel = model => { + const tmpHome = makeTempDir(); + const sessionId = `opus-rate-${process.pid}-${Date.now()}-${model}`; + const transcriptPath = path.join(tmpHome, 'session.jsonl'); + writeTranscript(transcriptPath, [ + { + type: 'assistant', + message: { + id: `msg_${model}`, + model, + usage: { input_tokens: 1_000_000, output_tokens: 1_000_000 }, + }, + }, + ]); + + try { + removeHarnessCostCache(sessionId); + const result = runScript( + { session_id: sessionId, transcript_path: transcriptPath }, + withTempHome(tmpHome) + ); + assert.strictEqual(result.code, 0, `Expected exit code 0, got ${result.code}`); + const metricsFile = path.join(tmpHome, '.claude', 'metrics', 'costs.jsonl'); + return JSON.parse(fs.readFileSync(metricsFile, 'utf8').trim()).estimated_cost_usd; + } finally { + removeHarnessCostCache(sessionId); + fs.rmSync(tmpHome, { recursive: true, force: true }); + } + }; + + assert.strictEqual( + priceModel('claude-opus-4-20250514'), + 90, + 'Expected dated Opus 4.0 to retain the legacy $15/$75 rate' + ); + assert.strictEqual( + priceModel('claude-opus-4-5-20251101'), + 30, + 'Expected Opus 4.5 to use the current $5/$25 rate' + ); + }) ? passed++ : failed++); + + // 11. Ignores stale harness-cost cache and falls back to transcript estimate (test('ignores stale harness-cost cache (>300s) and uses transcript estimate', () => { const tmpHome = makeTempDir(); const sessionId = 'harness-stale-' + Date.now(); diff --git a/tests/hooks/hooks.test.js b/tests/hooks/hooks.test.js index 4c15970b0..746aa88f7 100644 --- a/tests/hooks/hooks.test.js +++ b/tests/hooks/hooks.test.js @@ -2118,6 +2118,41 @@ async function runTests() { passed++; else failed++; + if ( + await asyncTest('keeps isMeta human prompts while filtering structured transcript noise', async () => { + const testDir = createTestDir(); + const transcriptPath = path.join(testDir, 'transcript.jsonl'); + const lines = [ + JSON.stringify({ type: 'user', isMeta: true, content: 'Prompt delivered by a channel plugin' }), + JSON.stringify({ type: 'user', isMeta: true, content: 'internal harness context' }), + JSON.stringify({ + type: 'user', + message: { role: 'user', content: [{ type: 'tool_result', content: 'tool output' }] }, + }), + ]; + fs.writeFileSync(transcriptPath, lines.join('\n')); + + const result = await runScript( + path.join(scriptsDir, 'session-end.js'), + JSON.stringify({ transcript_path: transcriptPath }), + { HOME: testDir, USERPROFILE: testDir } + ); + assert.strictEqual(result.code, 0); + + const sessionsDir = getCanonicalSessionsDir(testDir); + const sessionFiles = fs.readdirSync(sessionsDir).filter(file => file.endsWith('.tmp')); + assert.strictEqual(sessionFiles.length, 1, 'Should create one session file'); + const content = fs.readFileSync(path.join(sessionsDir, sessionFiles[0]), 'utf8'); + assert.ok(content.includes('Prompt delivered by a channel plugin')); + assert.ok(!content.includes('internal harness context')); + assert.ok(!content.includes('tool output')); + assert.ok(content.includes('Total user messages: 1')); + cleanupTestDir(testDir); + }) + ) + passed++; + else failed++; + if ( await asyncTest('extracts tool names and file paths from transcript', async () => { const testDir = createTestDir(); diff --git a/tests/lib/cost-estimate.test.js b/tests/lib/cost-estimate.test.js deleted file mode 100644 index bcb5906bc..000000000 --- a/tests/lib/cost-estimate.test.js +++ /dev/null @@ -1,114 +0,0 @@ -/** - * Tests for scripts/lib/cost-estimate.js - * - * Run with: node tests/lib/cost-estimate.test.js - */ - -const assert = require('assert'); - -const { estimateCost, RATE_TABLE } = require('../../scripts/lib/cost-estimate'); - -// Test helper -function test(name, fn) { - try { - fn(); - console.log(` \u2713 ${name}`); - return true; - } catch (err) { - console.log(` \u2717 ${name}`); - console.log(` Error: ${err.message}`); - return false; - } -} - -function runTests() { - console.log('\n=== Testing cost-estimate.js ===\n'); - - let passed = 0; - let failed = 0; - - // RATE_TABLE structure - console.log('RATE_TABLE:'); - - if ( - test('RATE_TABLE has haiku, sonnet, opus keys', () => { - assert.ok(RATE_TABLE.haiku, 'Missing haiku'); - assert.ok(RATE_TABLE.sonnet, 'Missing sonnet'); - assert.ok(RATE_TABLE.opus, 'Missing opus'); - assert.strictEqual(typeof RATE_TABLE.haiku.in, 'number'); - assert.strictEqual(typeof RATE_TABLE.haiku.out, 'number'); - assert.strictEqual(typeof RATE_TABLE.sonnet.in, 'number'); - assert.strictEqual(typeof RATE_TABLE.sonnet.out, 'number'); - assert.strictEqual(typeof RATE_TABLE.opus.in, 'number'); - assert.strictEqual(typeof RATE_TABLE.opus.out, 'number'); - }) - ) - passed++; - else failed++; - - // estimateCost tests - console.log('\nestimateCost:'); - - if ( - test('opus 1M/1M tokens returns 90', () => { - const cost = estimateCost('opus', 1_000_000, 1_000_000); - assert.strictEqual(cost, 90); - }) - ) - passed++; - else failed++; - - if ( - test('sonnet 1M/1M tokens returns 18', () => { - const cost = estimateCost('sonnet', 1_000_000, 1_000_000); - assert.strictEqual(cost, 18); - }) - ) - passed++; - else failed++; - - if ( - test('haiku 1M/1M tokens returns 4.8', () => { - const cost = estimateCost('haiku', 1_000_000, 1_000_000); - assert.strictEqual(cost, 4.8); - }) - ) - passed++; - else failed++; - - if ( - test('null model with 0 tokens returns 0', () => { - const cost = estimateCost(null, 0, 0); - assert.strictEqual(cost, 0); - }) - ) - passed++; - else failed++; - - if ( - test('full model name claude-opus-4-6 uses opus rates', () => { - const cost = estimateCost('claude-opus-4-6', 500, 200); - // (500 / 1_000_000) * 15 + (200 / 1_000_000) * 75 = 0.0075 + 0.015 = 0.0225 - const expected = Math.round(0.0225 * 1e6) / 1e6; - assert.strictEqual(cost, expected); - }) - ) - passed++; - else failed++; - - if ( - test('unknown model falls back to sonnet rates', () => { - const cost = estimateCost('unknown-model', 1_000_000, 1_000_000); - assert.strictEqual(cost, 18); - }) - ) - passed++; - else failed++; - - // Summary - console.log(`\nResults: ${passed} passed, ${failed} failed\n`); - return { passed, failed }; -} - -const { failed } = runTests(); -process.exit(failed > 0 ? 1 : 0);