Open Aura

cp .env.example .env # then set OPENAI_API_KEY to your OpenCode key docker compose up # Postgres + backend (MCP :8090) + viewer (:3000) docker compose up -d --build # rebuild images after pulling changes

Point your agent's MCP config at the local server (no auth in local mode):

// Claude Code / Cursor MCP config { "mcpServers": { "aura": { "url": "http://localhost:8090/mcp" } } }

Then, after a piece of work, ask your agent to"score this session with Aura."

The stackbakes the source into its images at build time(aura-appfrom.,aura-uivianext build) — there are no live source mounts. So after yougit pullnew code, a plaindocker compose upkeeps running theold cached imagesand you won't see the changes. Rebuild:

If a change still isn't showing (a stale layer cache), force a clean rebuild:

docker compose build --no-cache aura-app aura-ui docker compose up -d

Then hard-refresh the viewer (Ctrl/Cmd+Shift+R). The backend startup log prints the running version + build (e.g.Open Aura v0.2.0 · build …) so you can confirm the new code is live.

A referenceless read of yourprocess— no rubric, no test cases — across dimensions like prompting effectiveness, AI collaboration/steering, problem decomposition, and human contribution vs. AI reliance, plus a session archetype, insight cards, and measured telemetry (tokens, tools, "ships-it" lifecycle). Tools exposed:score_this_session,import_history,get_my_profile,whoami.

The redaction contract is the whole point:raw transcripts and file contents never leave your machine.The agent sends only truncated text excerpts and filepaths/metadata. In local mode the scoring and profile paths makeno outbound calls— everything stays in your local Postgres. (The only exception is the viewer's optionalLeaderboardtab, which does a plain read-onlyGETof the public leaderboard from vibelevel.ai — it sends none of your data.)

Anopt-in, off-by-defaultcapability that turns a scored session into concrete, tool-aware tips. The graph already maps the tools builders actually use —Claude skills, LangChain, the OpenAI SDK, MCP servers, and more — so Open Aura can look at what you built and nudge you toward them:"you hand-rolled this, but there's an established SDK for it,"or"there's a Claude skill for exactly this."Thesecheck-tipsare advisory only — they never change your Aura score.

It works preciselybecauseOpen Aura is local — it reads yourrealgit diffand full transcript (passed aslocal_context, never persisted or scored) instead of a redacted summary.

- src/aura_mcp/— the MCP server (server.py) + the single-local-user binding (local_auth.py).
- src/services/aura/— the referenceless scorer, signal extraction, archetypes, profile aggregation, and the scoringrubric(aura_model_coding.py/aura_model_writing.py) — open for contributions.
- src/core/— slim config, Postgres pool, and LLM provider routing.
- aura_mcp_app.py— the entrypoint that mounts the MCP app + the REST API +/health.
- packages/aura-ui/+web/— the React viewer components and the Next.js host that renders them (theaura-uicontainer).

This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.

Create crafted UI components inspired by the best 21st.dev design engineers.

Bring agent evaluations, observability, and synthetic test set generation directly into your IDE for free with Galileo's new MCP server

An MCP server to help AI assistants to answer questions and generate AccelByte Extend SDK code more effectively .

MCP server for AI Diagram Maker — generate beautiful software engineering diagrams directly inside Cursor, Claude Desktop, Claude Code, or any MCP-compatible AI agent

ALAPI MCP Tools,Call hundreds of API interfaces via MCP

AI-powered SVG animation generator that transforms static files into animated SVG components using the Allyson platform

MCP server that gives AI assistants on-demand access to 1,500+ amCharts docs, ~300 code examples, and 1000+ class API references.

APIMatic MCP Server is used to validate OpenAPI specifications using APIMatic. The server processes OpenAPI files and returns validation summaries by leveraging APIMatic’s API.

One shared context layer for AI agents and humans — live API specs, DB schemas, and versioned contracts across repos so every agent and teammate works from the same source of truth.