@hanfour.huang/caliber 0.1.2

package/README.md

@@ -0,0 +1,667 @@
+# Caliber
+
+**Measure the caliber of your AI-assisted engineering.** A self-hostable gateway, audit log, and evaluator for teams that want to know exactly what their AI coding assistants are doing — and how well.
+
+**精準衡量你的 AI 工程力。** 自架的 gateway / 稽核 / 評核平台，讓團隊清楚知道 AI 助理到底在做什麼、做得多好。
+
+---
+
+## Why / 為什麼需要這個工具
+
+Engineering managers need evidence-based data to evaluate how effectively their team uses AI coding assistants. Manual review of hundreds of AI sessions is impractical. This tool automates the process by:
+
+研發經理需要基於證據的資料來評估團隊使用 AI 程式助手的成效。手動審查數百個 AI 工作階段不切實際。本工具透過以下方式自動化此流程：
+
+1. **Extracting** usage data from local Claude Code (`~/.claude/`) and Codex (`~/.codex/`) storage
+2. **Analyzing** session patterns for decision-making quality and risk identification
+3. **Scoring** against a configurable evaluation standard (default: OneAD R&D standard)
+4. **Generating** structured reports with evidence and score recommendations
+
+---
+
+## Features / 功能特色
+
+- Reads Claude Code session metadata, facets, SQLite cost data, and JSONL conversations
+- Reads Codex SQLite thread data (tokens, models, sessions)
+- Detects decision-making patterns (iterative refinement, multi-task coordination, active corrections)
+- Detects risk identification signals (security awareness, performance discussions, bug catching)
+- Configurable evaluation standard — bring your own criteria, keywords, and thresholds
+- Multiple output formats: terminal (colored), JSON, Markdown, HTML
+- JSON output is machine-parseable (`--format json` emits clean JSON to stdout, progress logs go to stderr)
+- Noise filtering to exclude system messages and code review templates from analysis
+- `init-standard` command to export the default standard as a customization template
+- Data quality warnings when data sources are missing or incomplete
+
+---
+
+## Platform mode / 平台模式
+
+Starting with **v0.2.0** Caliber also ships as a self-hostable web platform with
+organization-scoped RBAC, invites, and an audit log. Use this mode if you want
+a shared workspace for a team rather than a per-engineer CLI report.
+
+**v0.3.0** adds an opt-in **gateway** that proxies Anthropic-native
+(`/v1/messages`) and OpenAI-compatible (`/v1/chat/completions`) traffic
+through a shared pool of upstream accounts:
+
+- Admins donate `sk-ant-...` API keys or OAuth bundles extracted from Claude
+  Code; the gateway's scheduler picks one per request based on priority,
+  concurrency, and rate-limit state.
+- Each user self-issues or receives an admin-issued platform API key
+  (`ak_...`) that authenticates against the gateway.
+- Usage + cost (per Anthropic/OpenAI token pricing) lands in a `usage_logs`
+  table, surfaced via per-user and per-org dashboards.
+
+**v0.4.0** adds an opt-in **evaluator** subsystem for performance evaluation
+(gated behind `ENABLE_EVALUATOR` feature flag):
+
+- **Content capture opt-in** — organization-level toggle; members see their
+  captured usage on `/dashboard/profile/evaluation`. 90-day default retention
+  (per-org override: 30/60/90). AES-256-GCM encryption with domain-separated
+  HKDF keys.
+- **Dual-layer evaluation** — rule-based scoring (always-on) + optional LLM
+  Deep Analysis (per-org opt-in). Costs dogfooded via self-gateway loopback
+  and tracked in `usage_logs`.
+- **Admin-customizable scoring rubrics** — platform defaults seeded for
+  English, Traditional Chinese, and Japanese; organizations can define custom
+  rubrics with dry-run preview. Zod-validated signal discriminated union
+  (keywords, thresholds, refusal rates, client mix, model diversity, cache
+  patterns, extended thinking, tool variety, iteration counts) — **extended
+  in v0.5.0** with six facet-based signal types.
+- **GDPR member-initiated delete request workflow** — members request deletion,
+  org admins approve (or auto-reject after 30 days). Retention purge and GDPR
+  execution run on separate cron workers.
+- **Labor-law-friendly transparency** — members always see their own full
+  evaluation report; team managers see redacted team views (LLM analysis
+  fields nulled unless they are also org admins). Leaderboard visibility is
+  opt-in per organization (privacy default).
+
+**v0.5.0** extends the v0.4.0 evaluator with **per-org LLM cost budgeting**
+and **opt-in LLM facet extraction** (gated behind `ENABLE_FACET_EXTRACTION`
+env + per-org `llm_facet_enabled`). All v0.4.0 behaviour preserved when both
+flags are off.
+
+- **Cost budget infrastructure** — every org gets `llm_monthly_budget_usd`
+  + `llm_budget_overage_behavior` (`degrade` skips over-budget calls,
+  `halt` stops all LLM until next UTC month). Spend tracked per-call in a
+  new `llm_usage_events` ledger, summed per UTC month, enforced before each
+  LLM call. Cost dashboard at `/dashboard/organizations/<id>/evaluator/costs`
+  with breakdowns by task / model / 6-month history; compact widget on the
+  evaluator status page.
+- **LLM facet extraction** — opt-in second LLM pass per session that
+  classifies each evaluation window's sessions into structured JSON
+  (`{sessionType, outcome, claudeHelpfulness, frictionCount, bugsCaughtCount,
+  codexErrorsCount}`). Extracted rows persisted to `request_body_facets`
+  table with `prompt_version` cache so the same LLM call doesn't fire twice.
+  Deterministic failures (parse / validation / timeout) write an error row
+  so they don't retry; transient failures (5xx, budget hit) skip silently.
+- **Six new rubric signal types** consume the facet aggregate:
+  `facet_claude_helpfulness`, `facet_friction_per_session`,
+  `facet_bugs_caught`, `facet_codex_errors`, `facet_outcome_success_rate`,
+  `facet_session_type_ratio`. Custom rubrics can opt in today; the rubric
+  editor ships an in-form Signal types reference.
+- **Platform default rubrics bumped to v1.1.0** — strictly additive: each
+  section gains one facet support (`facet_outcome_success` to `interaction`,
+  `facet_bugs_caught` to `riskControl`). Orgs without facet extraction see
+  zero scoring change.
+- **Report-page facet drill-down** — when facet rows exist for the period,
+  the user's evaluation report shows session-type distribution, success
+  rate, avg helpfulness, and bug/friction/codex counters. Hidden silently
+  when no rows exist.
+- **Observability artifacts** shipped under `ops/`: 3 Grafana dashboards
+  (evaluator / body-capture / GDPR), 11 Prometheus alert rules, 9 runbooks
+  in `docs/runbooks/`, and a post-release smoke workflow that auto-creates a
+  `release-blocker` issue when the canary fails.
+
+See [`docs/UPGRADE-v0.5.0.md`](docs/UPGRADE-v0.5.0.md) for the upgrade
+playbook (migrations 0004-0007, env flags, three-tier rollback).
+
+Quick start:
+```sh
+cd docker
+cp .env.example .env   # fill in OAuth + bootstrap email (+ gateway secrets if enabling)
+docker compose up -d                      # api + web + postgres + redis
+docker compose --profile gateway up -d    # opt-in: add gateway service
+```
+
+Images are published on every `v*` tag to:
+
+| Image | amd64 | arm64 |
+|-------|-------|-------|
+| `ghcr.io/hanfour/caliber-api` | ✅ | ✅ |
+| `ghcr.io/hanfour/caliber-gateway` (new in v0.3.0) | ✅ | ✅ |
+| `ghcr.io/hanfour/caliber-web` | ✅ | ❌ (dropped in v0.5.0; QEMU cross-build was unstable) |
+
+Operator guides:
+
+- **Try locally first**: [`docs/LOCAL_DEPLOY.md`](docs/LOCAL_DEPLOY.md) — 5-min path on your laptop, escalates to on-prem production
+- Self-hosting bring-up (api + web + gateway): [`docs/SELF_HOSTING.md`](docs/SELF_HOSTING.md)
+- Gateway operator + user reference: [`docs/GATEWAY.md`](docs/GATEWAY.md)
+
+**Cloud deploy templates** (alternatives to docker-compose self-hosting):
+
+- [Render Blueprint](deploy/render/README.md) — closest thing to one-click; provisions Postgres + 3 services; needs Upstash Redis externally
+- [Fly.io](deploy/fly/README.md) — three apps + Fly Postgres + Upstash; geographically distributed if you want
+- [Railway](deploy/railway/README.md) — native Postgres + Redis plugins; manual service creation per the README
+
+⚠️ Vercel is **not supported** — the gateway is a long-running Fastify
+server with BullMQ workers, doesn't fit Vercel's serverless model. See
+the deploy/ READMEs for what does work.
+
+CLI mode and platform mode share no runtime state; pick whichever fits.
+
+> **First time trying platform mode?** Start with
+> [`docs/GETTING_STARTED.md`](docs/GETTING_STARTED.md) — a 30-minute
+> end-to-end walkthrough that takes a fresh checkout to a working
+> personal AI gateway sharing your Claude.ai Pro/Max subscription
+> across all your devices.
+
+---
+
+## Data Sources / 資料來源
+
+| Source | Path | Data |
+|--------|------|------|
+| Claude Code Session Meta | `~/.claude/usage-data/session-meta/*.json` | Tokens, duration, tools, languages, git commits, first prompt |
+| Claude Code Facets | `~/.claude/usage-data/facets/*.json` | AI-generated session analysis: goals, outcomes, friction, helpfulness |
+| Claude Code SQLite | `~/.claude/__store.db` | Per-message cost (USD), model, duration |
+| Claude Code JSONL | `~/.claude/projects/*/*.jsonl` | Full conversation content for keyword signal scanning |
+| Codex SQLite | `~/.codex/state_5.sqlite` | Threads: tokens_used, model, title, git info |
+| Codex History | `~/.codex/history.jsonl` | Full user prompts by thread/session |
+| Codex Logs | `~/.codex/logs_2.sqlite` | Thread-level tool calls and error events |
+
+All data is read **locally and read-only**. No data is sent to any external service.
+
+所有資料皆為**本地端唯讀存取**，不會傳送至任何外部服務。
+
+---
+
+## Prerequisites / 系統需求
+
+- **Node.js** >= 18
+- **npm** (included with Node.js)
+- `~/.claude/` directory (from Claude Code usage)
+- `~/.codex/` directory (from Codex CLI usage, optional)
+
+---
+
+## Installation / 安裝
+
+### Recommended: Install from npm / 建議：從 npm 安裝
+
+```bash
+npm install -g @hanfour.huang/caliber
+
+# Verify installation
+caliber --version
+```
+
+### Update / 更新
+
+```bash
+npm install -g @hanfour.huang/caliber@latest
+```
+
+Caliber uses `~/.caliber.json` for CLI settings. On first run after upgrading
+from the older `aide` CLI, an existing `~/.aide.json` file is read, migrated to
+`~/.caliber.json`, and reported with a one-time deprecation notice.
+
+### Existing local-clone users / 已使用 clone 安裝的使用者
+
+If you previously installed from a cloned repo or `npm link`, migrate to the npm package:
+
+```bash
+npm unlink -g aide 2>/dev/null || npm uninstall -g @hanfour.huang/aide
+npm install -g @hanfour.huang/caliber@latest
+```
+
+### Development mode / 開發模式
+
+```bash
+git clone https://github.com/hanfour/caliber.git ~/caliber
+cd ~/caliber
+npm install
+npx tsx src/cli.ts --help
+```
+
+---
+
+## Quick Start / 快速開始
+
+```bash
+# Quick usage summary (last 7 days)
+caliber summary
+
+# Full evaluation report (last 30 days, terminal output)
+caliber report
+
+# Save report as Markdown
+caliber report --format markdown --output report.md
+
+# Save report as HTML
+caliber report --format html --output report.html
+
+# Monthly KPI report
+caliber monthly
+```
+
+---
+
+## Usage / 使用方式
+
+### Quick Summary / 快速摘要
+
+```bash
+# Last 7 days (default)
+caliber summary
+
+# Custom date range
+caliber summary --since 2026-03-01 --until 2026-03-31
+```
+
+Output:
+
+```
+AI Dev Usage Summary
+Period: 2026-03-01 ~ 2026-03-31
+
+Claude Code
+  Sessions:    57
+  Tokens:      259,336
+  Duration:    15676 min
+  Active Days: 9
+
+Codex
+  Sessions:    1
+  Tokens:      368,930
+  Active Days: 1
+```
+
+### Full Evaluation Report / 完整評核報告
+
+```bash
+# Default: last 30 days, text format, built-in OneAD standard
+caliber report
+
+# Current calendar month
+caliber monthly
+
+# Previous full calendar month
+caliber monthly --previous
+
+# Current calendar quarter
+caliber quarterly
+
+# Previous full calendar quarter
+caliber quarterly --previous
+
+# Custom date range
+caliber report --since 2026-03-01 --until 2026-04-14
+
+# Output as Markdown file
+caliber report --format markdown --output report.md
+
+# Output as HTML file
+caliber report --format html --output report.html
+
+# Output as JSON (machine-parseable, clean stdout)
+caliber report --format json --output report.json
+
+# Pipe JSON for programmatic consumption
+caliber report --format json 2>/dev/null | jq '.sections[].score'
+
+# Use a custom evaluation standard
+caliber report --standard my-standard.json
+
+# Include engineer/department metadata in report
+caliber report --engineer "Jane Doe" --department "R&D"
+```
+
+> **Note:** When using `--format json`, progress and status messages are written to stderr.
+> stdout contains only the JSON report, making it safe to pipe to `jq` or other tools.
+
+### Using the compiled CLI / 使用編譯後的 CLI
+
+If you are developing locally and have run `npm run build`, you can use `node dist/cli.js`:
+
+```bash
+node dist/cli.js report --since 2026-03-01 --until 2026-03-31
+node dist/cli.js summary
+node dist/cli.js monthly --previous --format markdown --output march.md
+node dist/cli.js report --format html --output report.html
+```
+
+---
+
+## CLI Reference / 命令參考
+
+### `caliber report`
+
+Generate a full evaluation report.
+
+```
+Options:
+  -s, --since <date>       Start date, YYYY-MM-DD (default: 30 days ago)
+  -u, --until <date>       End date, YYYY-MM-DD (default: today)
+  -f, --format <format>    Output: text | json | markdown (default: text)
+  -o, --output <file>      Write report to file instead of stdout
+  --standard <path>        Path to custom evaluation standard JSON
+  --engineer <name>        Engineer name for report identification
+  --department <name>      Department name for report identification
+```
+
+### `caliber summary`
+
+Quick usage summary for a date range.
+
+```
+Options:
+  -s, --since <date>       Start date, YYYY-MM-DD (default: 7 days ago)
+  -u, --until <date>       End date, YYYY-MM-DD (default: today)
+```
+
+### `caliber monthly`
+
+Generate a monthly KPI report.
+
+```
+Options:
+  -f, --format <format>    Output: text | json | markdown (default: text)
+  -o, --output <file>      Write report to file instead of stdout
+  --standard <path>        Path to custom evaluation standard JSON
+  --previous               Use the previous full calendar month
+```
+
+### `caliber quarterly`
+
+Generate a quarterly KPI report.
+
+```
+Options:
+  -f, --format <format>    Output: text | json | markdown (default: text)
+  -o, --output <file>      Write report to file instead of stdout
+  --standard <path>        Path to custom evaluation standard JSON
+  --previous               Use the previous full calendar quarter
+```
+
+### `caliber init-standard`
+
+Export the default evaluation standard as a JSON template for customization.
+
+```
+Options:
+  -o, --output <file>      Output file path (default: eval-standard.json)
+```
+
+---
+
+## Report Structure / 報告結構
+
+The generated report contains the following sections:
+
+### 1. Management Summary / 管理摘要
+
+Management-facing overview for monthly/quarterly KPI review:
+
+- Overall headline
+- Period assessment
+- Key observations
+- Recommended follow-up actions
+
+### 2. Usage Overview / 使用概覽
+
+Quantitative metrics for both Claude Code and Codex:
+
+- Total sessions, tokens (input/output), estimated cost
+- Active days, duration
+- Top projects by token usage
+- Top tools used (Bash, Read, Edit, etc.)
+- Model breakdown
+
+### 3-N. Evaluation Sections / 評核區段
+
+Each section defined in the evaluation standard generates:
+
+- **Summary** — aggregate statistics
+- **Usage evidence** — workload/depth indicators such as sessions, tool usage, follow-up prompts
+- **Score evidence** — threshold-relevant evidence used for 100% / 120% scoring
+- **Evidence signals** — grouped by type (iterative refinement, bugs caught, security awareness, etc.)
+- **Metrics** — numeric indicators used for scoring
+
+### Final. Score Recommendation / 分值建議
+
+For each evaluation section:
+
+- **Score**: Standard (100%) or Superior (120%)
+- **Label**: Human-readable grade
+- **Reason**: Evidence-backed explanation referencing the criteria
+
+### Data Quality Warnings / 資料品質警告
+
+The report includes data quality warnings when:
+
+- Required data sources (`~/.claude/usage-data/session-meta`) are missing
+- Sessions exist but no facets are found (qualitative analysis limited)
+- No keyword signals detected (JSONL files may be missing)
+- No sessions found at all in the evaluation period
+
+---
+
+## Custom Evaluation Standards / 自訂評核標準
+
+The built-in default is the OneAD R&D AI-Application Evaluation Standard. To create your own:
+
+### Step 1: Export the default template / 匯出預設範本
+
+```bash
+npx tsx src/cli.ts init-standard --output my-standard.json
+```
+
+### Step 2: Edit the JSON file / 編輯 JSON 檔案
+
+Key fields you can customize:
+
+| Field | Purpose |
+|-------|---------|
+| `name` | Standard name shown in report header |
+| `sections[]` | Array of evaluation sections (add/remove/reorder) |
+| `sections[].id` | Unique section identifier |
+| `sections[].name` | Section display name |
+| `sections[].weight` | KPI weight (display only) |
+| `sections[].keywords` | Conversation scanning keywords |
+| `sections[].thresholds` | Numeric thresholds for Superior score |
+| `sections[].superiorRules` | Optional rule for combining thresholds |
+| `sections[].standard` | 100% score criteria text |
+| `sections[].superior` | 120% score criteria text |
+| `noiseFilters` | Rules to exclude system/template messages |
+
+### Step 3: Use it / 使用自訂標準
+
+```bash
+npx tsx src/cli.ts report --standard my-standard.json
+```
+
+### Example: Adding a new section / 新增評核區段範例
+
+```json
+{
+  "id": "collaboration",
+  "name": "AI-Human Collaboration Quality",
+  "weight": "30%",
+  "standard": {
+    "score": 100,
+    "label": "Standard",
+    "criteria": ["Uses AI for routine tasks", "Follows AI suggestions without modification"]
+  },
+  "superior": {
+    "score": 120,
+    "label": "Superior",
+    "criteria": ["Actively debates with AI on design decisions", "Synthesizes multiple AI suggestions into novel solutions"]
+  },
+  "keywords": ["design", "architecture", "trade-off", "pattern", "alternative"],
+  "thresholds": {
+    "iterativeRatio": 0.4,
+    "keywordHits": 15
+  },
+  "superiorRules": {
+    "mode": "grouped",
+    "strongThresholds": ["iterativeRatio", "keywordHits"],
+    "supportThresholds": ["avgToolUses"],
+    "minStrongMatched": 1,
+    "minSupportMatched": 0
+  }
+}
+```
+
+### Superior Rules / 升等規則
+
+`superiorRules.mode = "any"` — any matched threshold is enough for 120%.
+
+`superiorRules.mode = "grouped"` — separate strong evidence from support evidence. Strong evidence must meet a minimum count; support evidence alone is not sufficient.
+
+Keys referenced by `strongThresholds` and `supportThresholds` must also exist in `thresholds`.
+
+### Available threshold keys / 可用門檻鍵值
+
+| Key | Description |
+|-----|-------------|
+| `iterativeRatio` | Ratio of iterative/multi-task sessions to total |
+| `correctionCount` | Number of user corrections/interruptions |
+| `keywordHits` | Number of keyword signal matches |
+| `avgToolUses` | Average tool uses per session |
+| `securityCount` | Security-related keyword matches |
+| `performanceCount` | Performance-related keyword matches |
+| `bugsCaught` | AI-generated bugs caught (from facets) |
+| `frictionSessions` | Sessions with friction events |
+| `codexIterativeSessions` | Codex threads with strong iterative evidence |
+| `codexMultiTurnSessions` | Codex multi-turn threads |
+| `codexFollowUpCount` | Codex follow-up user prompts |
+| `codexDeepSessions` | Codex high-depth threads |
+| `codexErrorSessions` | Codex threads with logged errors |
+
+---
+
+## Default Evaluation Standard / 預設評核標準
+
+The built-in OneAD standard evaluates two dimensions:
+
+### AI Interaction & Decision (20% KPI weight) / AI 交互與決策
+
+| Grade | Criteria |
+|-------|----------|
+| **Standard (100%)** | Actively use AI for coding; clear decision notes |
+| **Superior (120%)** | Multi-iteration guidance (A->B->C); system-constraint-aware optimization |
+
+### AI Identification & Risk Control (50% KPI weight) / AI 識別與風險控管
+
+| Grade | Criteria |
+|-------|----------|
+| **Standard (100%)** | Catch common AI errors/hallucinations; stable code |
+| **Superior (120%)** | Identify critical risks (security, performance, memory); produce SOP/Wiki for team sharing |
+
+---
+
+## Architecture / 架構
+
+```
+src/
+├── cli.ts                    # CLI entry point (commander)
+├── types.ts                  # TypeScript type definitions
+├── standard.ts               # Load & validate evaluation standards
+├── period.ts                 # Date period resolution (monthly/quarterly)
+├── data-quality.ts           # Data source completeness checks
+├── utils.ts                  # Shared utilities (noise filter)
+├── extractors/
+│   ├── claude-code.ts        # Read ~/.claude/ data (JSONL, SQLite, JSON)
+│   └── codex.ts              # Read ~/.codex/ data (SQLite, JSONL)
+├── analyzers/
+│   ├── usage.ts              # Aggregate quantitative usage metrics
+│   └── section.ts            # Generic section analyzer (facets + keywords + thresholds)
+└── reporters/
+    └── report.ts             # Render reports (text, JSON, Markdown)
+
+templates/
+└── eval-standard.json        # Default OneAD evaluation standard (source of truth)
+
+tests/
+├── cli.test.ts               # CLI regression tests (subprocess)
+├── section.test.ts           # Section analyzer unit tests
+├── standard.test.ts          # Standard loader/validator tests
+├── data-quality.test.ts      # Data quality checker tests
+└── fixtures/                 # Test fixture files
+```
+
+### Pipeline / 處理流程
+
+```
+Extract --> Analyze --> Score --> Report
+
+1. Extract:  Read session-meta, facets, SQLite, JSONL from local stores
+2. Analyze:  Aggregate usage + run each section through generic analyzer
+3. Score:    Compare metrics against section thresholds
+4. Report:   Render in chosen format with evidence and recommendations
+```
+
+---
+
+## Development / 開發
+
+### Scripts / 腳本
+
+```bash
+npm run build        # Compile TypeScript to dist/
+npm run dev          # Run CLI directly via tsx (no build needed)
+npm run test         # Run test suite (vitest)
+npm run test:watch   # Run tests in watch mode
+```
+
+### Running tests / 執行測試
+
+```bash
+# Run all tests
+npm test
+
+# Run a specific test file
+npx vitest run tests/section.test.ts
+
+# Watch mode
+npm run test:watch
+```
+
+### Project conventions / 專案慣例
+
+- All progress/status messages are written to **stderr**; report output goes to **stdout**
+- JSON output (`--format json`) is guaranteed clean on stdout for piping
+- SQLite connections are wrapped in `try/finally` to prevent resource leaks
+- The evaluation standard template (`templates/eval-standard.json`) is the single source of truth
+- Custom standards inherit default `noiseFilters` if not specified
+
+---
+
+## Troubleshooting / 問題排除
+
+### No sessions found
+
+- Verify `~/.claude/usage-data/session-meta/` contains JSON files
+- Check the date range matches when the AI tools were used
+- For Codex, verify `~/.codex/state_5.sqlite` exists
+
+### Empty facets
+
+- Facets are generated asynchronously by Claude Code after sessions end
+- Recent sessions may not have facets yet
+- The tool will show a data quality warning in this case
+
+### JSON output contains extra text
+
+This was fixed in v0.1.0. All progress messages now go to stderr. If you encounter this, ensure you are using the latest version. Use `2>/dev/null` to suppress stderr when piping:
+
+```bash
+npx tsx src/cli.ts report --format json 2>/dev/null | jq .
+```
+
+---
+
+## License
+
+MIT

package/dist/analyzers/section.d.ts

@@ -0,0 +1,9 @@
+import type { ClaudeCodeSession, ClaudeCodeFacet, ClaudeCodeConversationSignal, CodexSession, CodexConversationSignal, CodexSessionInsight, EvalSectionDef, EvalSectionResult } from "../types.js";
+/**
+ * Generic section analyzer.
+ *
+ * Collects evidence from facets, sessions, and conversation signals,
+ * then determines standard vs superior score based on section thresholds.
+ */
+export declare function analyzeSection(section: EvalSectionDef, claudeSessions: ClaudeCodeSession[], facets: Map<string, ClaudeCodeFacet>, claudeSignals: ClaudeCodeConversationSignal[], codexSessions: CodexSession[], codexInsights: Map<string, CodexSessionInsight>, codexSignals: CodexConversationSignal[]): EvalSectionResult;
+//# sourceMappingURL=section.d.ts.map

package/dist/analyzers/section.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"section.d.ts","sourceRoot":"","sources":["../../src/analyzers/section.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EACV,iBAAiB,EACjB,eAAe,EACf,4BAA4B,EAC5B,YAAY,EACZ,uBAAuB,EACvB,mBAAmB,EACnB,cAAc,EACd,iBAAiB,EAClB,MAAM,aAAa,CAAC;AAsiBrB;;;;;GAKG;AACH,wBAAgB,cAAc,CAC5B,OAAO,EAAE,cAAc,EACvB,cAAc,EAAE,iBAAiB,EAAE,EACnC,MAAM,EAAE,GAAG,CAAC,MAAM,EAAE,eAAe,CAAC,EACpC,aAAa,EAAE,4BAA4B,EAAE,EAC7C,aAAa,EAAE,YAAY,EAAE,EAC7B,aAAa,EAAE,GAAG,CAAC,MAAM,EAAE,mBAAmB,CAAC,EAC/C,YAAY,EAAE,uBAAuB,EAAE,GACtC,iBAAiB,CAwKnB"}