@jeeva1398/eventa 0.1.0

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Jeeva1398
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/README.md

@@ -0,0 +1,53 @@
+# eventa
+
+**Offline AI assistant for Node.js developers.** It explains crashes, reviews your diffs and audits dependencies, using a small code model that runs on your own machine. No API key, and no code leaves your laptop.
+
+```bash
+npx @jeeva1398/eventa explain --run "node app.js"   # run it, explain the crash
+npx @jeeva1398/eventa review                        # review staged changes
+npx @jeeva1398/eventa deps                          # vulnerabilities, outdated, unused, missing
+```
+
+Or install once with `npm i -g @jeeva1398/eventa`, then run `eventa explain`, `eventa review`, and so on.
+
+Requires Node.js 20+. On first use Eventa installs a small llama.cpp runtime (~80 MB) and downloads the model (~1 GB, checksum-verified). If [Ollama](https://ollama.com) is running, Eventa uses it instead.
+
+## Commands
+
+### `eventa explain`
+Give it an error in any of these ways:
+- pipe it in: `node app.js 2>&1 | eventa explain`
+- read it from a log: `eventa explain --file crash.log`
+- run a command: `eventa explain --run "npm test"` or `eventa explain --run "npx tsc --noEmit"` for TypeScript compile errors
+- paste it as arguments
+
+Eventa parses the stack trace or `tsc` output, reads the source lines it points to in your project, and answers with **Cause**, **Fix** and **Prevent**. Known error codes come with built-in reference notes: Node (`ERR_REQUIRE_ESM`, `ECONNREFUSED`, …), TypeScript (`TS2345`, `TS18048`, …), Prisma (`P2002`, `P2025`, …) and NestJS dependency-injection errors.
+
+### `eventa review`
+Reviews `git diff --staged`, falling back to unstaged changes. Use `--base main` to review everything since `main`. Each file gets static checks first (missing `await`, SQL/command injection, `eval`, sync fs, empty `catch`, disabled TLS, hard-coded secrets), then the model confirms or dismisses them and reports `- [severity] line N: problem → fix`.
+
+### `eventa deps`
+Runs `npm audit` and `npm outdated` and scans your imports. It prints the vulnerabilities with their **exact fix commands**, outdated packages (flagging major bumps), unused dependencies and packages that are imported but not declared. Then it adds AI advice on upgrade risk. Use `--no-ai` for the report only.
+
+### More
+- `eventa ask <question>`: any Node.js question
+- `eventa doctor`: check Node, Ollama and the built-in runtime
+- `eventa model list | pull [id] | use <id> | rm <id>`: manage local models (`qwen2.5-coder-1.5b`, `eventa-1.5b`, or any `.gguf` URL or path)
+- `eventa config get | set <key> <value> | path`: settings in `~/.eventa/config.json`
+
+Global flags: `-p, --provider auto|ollama|local`, `-m, --model <name>`, `--host <url>`, `--raw`. `explain`, `review` and `deps` also accept `--json` for scripts and CI.
+
+| Setting | Default | Env |
+|---|---|---|
+| `provider` | `auto` | `EVENTA_PROVIDER` |
+| `model` (Ollama) | `hf.co/jeeva1398/eventa-1.5b-gguf` | `EVENTA_MODEL` |
+| `localModel` (built-in) | `eventa-1.5b` | `EVENTA_LOCAL_MODEL` |
+| `ollamaHost` | `http://localhost:11434` | `EVENTA_OLLAMA_HOST` / `OLLAMA_HOST` |
+| `contextSize` | `8192` | |
+| `temperature` | `0.2` | |
+
+## Limitations
+This is a 1.5B model. It is fast and private, but it can be wrong, so treat its answers as a strong hint, not a verdict. The facts Eventa computes itself (stack frames, audit fixes, unused/missing packages, static checks) are deterministic.
+
+## License
+MIT. Source and training pipeline: https://github.com/Jeeva1398/eventa