Atena-IT

Operate the Open Arena REST API: start the server and make authenticated requests.

Run the arena evaluation sweep and interpret last_run.tsv and the leaderboard matrix.

Start and operate the autonomous reward-R&D experiment loop (setup, confirm, then run until interrupted).

Add or iterate on a custom reward in src/rewards/ and wire it into the sweep.

Configure an Open Arena sweep: author/edit config.yaml datasets, rewards, metrics, and experiments blocks.

Prepare evaluation datasets: write prepare_data.py, wire loaders in config.yaml, smoke-test that rows load.

Query live LLM token pricing (including cache read/write rates), compute workload cost from known token counts, and set up webhook notifications for pricing changes. Use when the user asks what a model costs per 1M tokens, wants to compare model pricing across providers or currencies, needs total spend for a workload with known input/output (and optionally cache) token counts, asks what prompt caching costs or saves for a model, or wants alerts when model prices change or models are added, deprecated, or removed. This skill shells out to the tokenpricing CLI and the notifier service. Do not use it for counting tokens from raw text, subscription or invoice questions, or billing topics that are not model-pricing math. Triggers include "what does openai/gpt-5.2 cost", "compare Claude and GPT pricing in EUR", "what would 250000 input and 40000 output tokens cost", "how much do cache reads cost on claude", "would prompt caching save me money here", "notify my webhook when claude prices change", or "show me machine

AI-native project management CLI for GitHub. Use when the user wants to inspect or update the planning hierarchy (epics, sprints, user stories), drive story lifecycle transitions (start work, ready for review, close), or compile inbound notes, transcripts, uploads, and feature requests into structured wiki knowledge before changing the plan. Do not use for reading a single issue or PR without project-management intent; prefer standard gh commands for those one-off lookups. Triggers include "what's the current sprint status", "create a story for this feature", "move this story to in-progress", "close out the sprint", "update the plan based on today's meeting notes", or any request involving epics, sprints, story lifecycle, or wiki-backed planning updates.

Teach AI coding assistants how to install, verify, and operate the local dit-commandline CLI in this repository. Use when the task is to run `dit` commands, set up a Dit repository, inspect available CLI commands/options, or choose the right local invocation mode (`uv run dit`, `uv run --package dit-commandline dit`, or plain `dit`).