npm - @rely-ai/caliber - Versions diffs - 0.4.3 → 0.4.5 - Mend

@rely-ai/caliber 0.4.3 → 0.4.5

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Files changed (3) hide show

package/README.md CHANGED Viewed

@@ -1,155 +1,118 @@
 # Caliber
-Open-source CLI that analyzes your project and generates optimized configuration files for AI coding agents (Claude Code, Cursor). Use your **current seat** (Claude Code Pro/Max/Team or Cursor subscription, no API key) or bring your own API key — supports Claude Code CLI (`claude -p`), Cursor (ACP), Anthropic, OpenAI, Google Vertex AI, and any OpenAI-compatible endpoint.
+[![npm version](https://img.shields.io/npm/v/@rely-ai/caliber)](https://www.npmjs.com/package/@rely-ai/caliber)
+[![license](https://img.shields.io/npm/l/@rely-ai/caliber)](./LICENSE)
+[![node](https://img.shields.io/node/v/@rely-ai/caliber)](https://nodejs.org)
-## Installation
+**Analyze your codebase. Generate optimized AI agent configs. One command.**
-```bash
-npm install -g @rely-ai/caliber
-```
+Caliber scans your project — languages, frameworks, dependencies, file structure — and generates tailored config files for Claude Code and Cursor. If configs already exist, it audits them and suggests improvements.
+**No API key required** — use your existing Claude Code or Cursor subscription. Or bring your own key (Anthropic, OpenAI, Vertex AI, any OpenAI-compatible endpoint).
 ## Quick Start
 ```bash
-# Option 1: Use your current seat — no API key
-caliber config   # choose "Claude Code" (Pro/Max/Team) or "Cursor"
-# Or: export CALIBER_USE_CLAUDE_CLI=1  or  CALIBER_USE_CURSOR_SEAT=1
-# Option 2: Set an API key
-export ANTHROPIC_API_KEY=sk-ant-...
-# Then run
-caliber init
+npx @rely-ai/caliber init
 ```
-## What It Does
-Caliber scans your codebase — languages, frameworks, file structure, existing configs — and generates tailored configuration files:
-- **CLAUDE.md** — Project context for Claude Code (commands, architecture, conventions)
-- **.cursorrules** / **.cursor/rules/** — Rules for Cursor
-- **Skills** — Reusable skill files following the [OpenSkills](https://agentskills.io) standard
-If you already have these files, Caliber audits them against your actual codebase and suggests targeted improvements — keeping what works, fixing what's stale, adding what's missing.
-## Commands
-| Command | Description |
-|---------|-------------|
-| `caliber init` | Scan project and generate agent config |
-| `caliber update` | Re-analyze and regenerate (alias: `regenerate`, `regen`) |
-| `caliber config` | Configure LLM: Cursor (your seat), API key, and model |
-| `caliber refresh` | Update docs based on recent git changes |
-| `caliber score` | Score your config quality (deterministic, no LLM). Supports `--agent claude\|cursor\|both` |
-| `caliber recommend` | Discover skills from [skills.sh](https://skills.sh) |
-| `caliber undo` | Revert all changes made by Caliber |
-| `caliber status` | Show current setup status |
-| `caliber hooks install` | Install Claude Code auto-refresh hook |
-| `caliber hooks remove` | Remove Claude Code auto-refresh hook |
-| `caliber hooks install-precommit` | Install git pre-commit hook for auto-refresh |
-| `caliber hooks remove-precommit` | Remove git pre-commit hook |
-| `caliber hooks status` | Show installed hooks |
-| `caliber learn install` | Install session learning hooks |
-| `caliber learn status` | Show learned insights from sessions |
-| `caliber learn observe` | Manually feed a tool event for analysis |
-| `caliber learn finalize` | Analyze captured events and extract patterns |
-| `caliber learn remove` | Remove learning hooks |
-## Supported LLM Providers
-| Provider | How to use | Notes |
-|----------|------------|-------|
-| **Claude Code (current seat)** | `caliber config` → "Claude Code", or `CALIBER_USE_CLAUDE_CLI=1` | Uses your Pro/Max/Team login via `claude -p`. No API key; install [Claude Code CLI](https://claude.ai/install) and run `claude` once to log in. |
-| **Cursor (current seat)** | `caliber config` → "Cursor", or `CALIBER_USE_CURSOR_SEAT=1` | Uses your Cursor subscription via [Cursor Agent (ACP)](https://cursor.com/docs/cli/acp). No API key; run `agent login` once if needed. |
-| **Anthropic (Claude)** | `ANTHROPIC_API_KEY` | Claude Sonnet 4.6 default. Get an API key at [console.anthropic.com](https://console.anthropic.com) (same company as Claude Pro/Team/Max; API is separate billing). |
-| **Google Vertex AI** | `VERTEX_PROJECT_ID` or `GCP_PROJECT_ID` | Uses ADC by default. Region `us-east5`. Set `VERTEX_REGION`, `VERTEX_SA_CREDENTIALS` as needed. |
-| **OpenAI** | `OPENAI_API_KEY` | GPT-4.1 default. |
-| **Custom endpoint** | `OPENAI_API_KEY` + `OPENAI_BASE_URL` | Any OpenAI-compatible API (Ollama, vLLM, Together, etc.) |
-Override the model with `CALIBER_MODEL=<model-name>` or via `caliber config`.
+That's it. On first run, Caliber walks you through provider setup interactively.
-### Vertex AI Setup
+Or install globally:
 ```bash
-# Minimal — uses gcloud ADC and defaults
-export VERTEX_PROJECT_ID=my-gcp-project
+npm install -g @rely-ai/caliber
 caliber init
+```
-# With custom region
-export VERTEX_PROJECT_ID=my-gcp-project
-export VERTEX_REGION=europe-west1
-caliber init
+> **Already have an API key?** Skip the interactive setup:
+> ```bash
+> export ANTHROPIC_API_KEY=sk-ant-...
+> npx @rely-ai/caliber init
+> ```
-# With service account credentials (inline JSON)
-export VERTEX_PROJECT_ID=my-gcp-project
-export VERTEX_SA_CREDENTIALS='{"type":"service_account",...}'
-caliber init
+## How It Works
-# With service account credentials (file path via GOOGLE_APPLICATION_CREDENTIALS)
-export VERTEX_PROJECT_ID=my-gcp-project
-export GOOGLE_APPLICATION_CREDENTIALS=/path/to/service-account.json
+```
 caliber init
+│
+├─ 1. Scan        Analyze languages, frameworks, dependencies, file structure,
+│                  and existing agent configs in your project
+│
+├─ 2. Generate    LLM creates tailored config files based on your codebase
+│                  (or audits existing ones and suggests improvements)
+│
+├─ 3. Review      You see a diff of proposed changes — accept, refine via
+│                  chat, or decline
+│
+└─ 4. Apply       Files are written with automatic backups, before/after
+                   score is displayed
 ```
-## Flow: install and init
+### What It Generates
-- **Install:** `npm install -g @caliber-ai/caliber` runs the postinstall script and prints the "Get started" message. With **npx** (`npx @caliber-ai/caliber init`), there is no install step; the binary runs from cache.
-- **Config:** User runs `caliber config` (or sets env vars). Config is stored in `~/.caliber/config.json` or taken from `ANTHROPIC_API_KEY` / `OPENAI_API_KEY` / `VERTEX_PROJECT_ID` / `CALIBER_USE_CLAUDE_CLI` / `CALIBER_USE_CURSOR_SEAT`.
-- **Init:** `caliber init` calls `loadConfig()`: env vars override; if none, it reads `~/.caliber/config.json`. If no config, init prints the options and exits. Otherwise it scans the project, generates config with the chosen LLM (**Claude Code CLI**, Cursor ACP, Anthropic, OpenAI, or Vertex), then review and apply.
+| File | Purpose |
+|------|---------|
+| `CLAUDE.md` | Project context for Claude Code — commands, architecture, conventions |
+| `.cursorrules` / `.cursor/rules/` | Rules for Cursor |
+| Skills (`.claude/skills/`, `.cursor/skills/`) | Reusable skill files following the [OpenSkills](https://agentskills.io) standard |
+| `AGENTS.md` | Agent collaboration guide |
-See [docs/FLOW.md](docs/FLOW.md) for the full step-by-step.
+If these files already exist, Caliber audits them against your actual codebase and suggests targeted improvements — keeping what works, fixing what's stale, adding what's missing.
-## How It Works
-1. **Scan** — Analyzes your code, dependencies, file structure, and existing agent configs
-2. **Generate** — LLM creates config files tailored to your project
-3. **Review** — You accept, refine via chat, or decline the proposed changes
-4. **Apply** — Config files are written to your project with backups, and a before/after score is displayed
-Caliber also auto-generates `AGENTS.md` and configures `.claude/settings.json` permissions during init.
-### Scoring
+## Commands
-Caliber includes a deterministic scoring system (no LLM needed) that evaluates your agent config across 6 categories: existence, quality, coverage, accuracy, freshness, and bonus. Scoring is target-aware — it only checks what's relevant to your chosen platform:
+| Command | Description |
+|---------|-------------|
+| `caliber init` | Scan project, generate/audit agent configs, review and apply |
+| `caliber score` | Score your config quality (deterministic, no LLM needed) |
+| `caliber recommend` | Discover and install skills from [skills.sh](https://skills.sh) |
+| `caliber config` | Configure LLM provider, API key, and model |
 ```bash
-caliber score               # Auto-detect target from existing files
-caliber score --agent claude  # Score for Claude Code only
-caliber score --agent both    # Score for Claude Code + Cursor
+caliber init --agent claude      # Target Claude Code only
+caliber init --agent cursor      # Target Cursor only
+caliber init --agent both        # Target both
+caliber init --dry-run           # Preview without writing files
+caliber score --json             # Machine-readable output
 ```
-During `caliber init`, a before/after score is displayed so you can see the improvement.
+## LLM Providers
-### Auto-refresh
+| Provider | Setup | Notes |
+|----------|-------|-------|
+| **Claude Code** (your seat) | `caliber config` → Claude Code | No API key. Uses your Pro/Max/Team login via `claude -p`. |
+| **Cursor** (your seat) | `caliber config` → Cursor | No API key. Uses your subscription via Cursor Agent (ACP). |
+| **Anthropic** | `export ANTHROPIC_API_KEY=sk-ant-...` | Claude Sonnet 4.6 default. [Get key](https://console.anthropic.com). |
+| **OpenAI** | `export OPENAI_API_KEY=sk-...` | GPT-4.1 default. |
+| **Vertex AI** | `export VERTEX_PROJECT_ID=my-project` | Uses ADC. Region `us-east5`. |
+| **Custom endpoint** | `OPENAI_API_KEY` + `OPENAI_BASE_URL` | Any OpenAI-compatible API (Ollama, vLLM, Together, etc.) |
-During `caliber init`, you'll be prompted to choose how docs auto-refresh:
+Override the model for any provider: `export CALIBER_MODEL=<model-name>` or use `caliber config`.
-- **Claude Code hook** — refreshes docs when Claude Code sessions end
-- **Git pre-commit hook** — refreshes docs before each commit
-- **Both** — enables both hooks
-- **Skip** — install later with `caliber hooks install` or `caliber hooks install-precommit`
+<details>
+<summary>Vertex AI advanced setup</summary>
 ```bash
-caliber hooks install              # Install Claude Code hook
-caliber hooks install-precommit    # Install git pre-commit hook
-caliber hooks remove               # Remove Claude Code hook
-caliber hooks remove-precommit     # Remove pre-commit hook
-caliber hooks status               # Show installed hooks
-```
-### Session Learning
+# Custom region
+export VERTEX_PROJECT_ID=my-gcp-project
+export VERTEX_REGION=europe-west1
-Caliber can observe your Claude Code sessions and extract reusable instructions:
+# Service account credentials (inline JSON)
+export VERTEX_PROJECT_ID=my-gcp-project
+export VERTEX_SA_CREDENTIALS='{"type":"service_account",...}'
-```bash
-caliber learn install    # Install learning hooks
-caliber learn status     # Check what's been captured
+# Service account credentials (file path)
+export VERTEX_PROJECT_ID=my-gcp-project
+export GOOGLE_APPLICATION_CREDENTIALS=/path/to/service-account.json
 ```
+</details>
 ## Requirements
 - Node.js >= 20
-- An LLM: use your **Claude Code** or **Cursor** subscription (run `caliber config` → Claude Code or Cursor; for Claude Code run `claude` once to log in), or set an API key for Anthropic, OpenAI, or Vertex
+- One LLM provider: your **Claude Code** or **Cursor** subscription (no API key), or an API key for Anthropic / OpenAI / Vertex AI
 ## Contributing
@@ -162,7 +125,7 @@ npm run test     # Run tests
 npm run build    # Compile
 ```
-This project uses [conventional commits](https://www.conventionalcommits.org/) — `feat:` for features, `fix:` for bug fixes. See the [CLAUDE.md](./CLAUDE.md) for architecture details.
+Uses [conventional commits](https://www.conventionalcommits.org/) — `feat:` for features, `fix:` for bug fixes.
 ## License

package/dist/bin.js CHANGED Viewed

@@ -2744,9 +2744,22 @@ function extractNpmDeps(dir) {
     "rimraf",
     "cross-env",
     "dotenv",
-    "nodemon"
+    "nodemon",
+    "husky",
+    "lint-staged",
+    "commitlint",
+    "@commitlint/cli",
+    "@commitlint/config-conventional"
   ]);
-  return Object.keys(deps).filter((d) => !trivial.has(d) && !d.startsWith("@types/")).slice(0, 30);
+  const trivialPatterns = [
+    /^@rely-ai\//,
+    /^@caliber-ai\//,
+    /^eslint-/,
+    /^@eslint\//,
+    /^prettier-/,
+    /^@typescript-eslint\//
+  ];
+  return Object.keys(deps).filter((d) => !trivial.has(d) && !d.startsWith("@types/") && !trivialPatterns.some((p) => p.test(d))).slice(0, 30);
 }
 function extractPythonDeps(dir) {
   const reqTxt = readFileOrNull(join(dir, "requirements.txt"));
@@ -3834,6 +3847,7 @@ async function initCommand(options) {
 `));
   const targetAgent = options.agent || await promptAgent();
   const baselineScore = computeLocalScore(process.cwd(), targetAgent);
+  console.log(chalk4.hex("#6366f1").bold("  Current project score\n"));
   displayScore(baselineScore);
   const hasExistingConfig = !!(fingerprint.existingConfigs.claudeMd || fingerprint.existingConfigs.claudeSettings || fingerprint.existingConfigs.claudeSkills?.length || fingerprint.existingConfigs.cursorrules || fingerprint.existingConfigs.cursorRules?.length);
   if (hasExistingConfig && baselineScore.score === 100) {

package/package.json CHANGED Viewed

@@ -1,7 +1,7 @@
 {
   "name": "@rely-ai/caliber",
-  "version": "0.4.3",
-  "description": "Open-source CLI for configuring coding agent environments (CLAUDE.md, .cursorrules, skills). Bring your own LLM.",
+  "version": "0.4.5",
+  "description": "Analyze your codebase and generate optimized AI agent configs (CLAUDE.md, .cursorrules, skills) — no API key needed",
   "type": "module",
   "bin": {
     "caliber": "./dist/bin.js"
@@ -34,16 +34,16 @@
   "license": "MIT",
   "author": "Rely AI",
   "keywords": [
-    "caliber",
-    "cli",
-    "coding-agent",
-    "claude",
+    "ai-agent",
     "claude-code",
-    "cursor",
-    "ai",
-    "developer-tools",
     "claude-md",
+    "cursor",
     "cursorrules",
+    "coding-agent",
+    "developer-tools",
+    "ai-config",
+    "skills",
+    "cli",
     "open-source"
   ],
   "engines": {