python-ai
SkillAI & modelsBuild and troubleshoot Python AI applications involving model APIs, local LLMs, retrieval, tools, agents, prompts, evaluation, or AI dependencies.
Available today. Use it from your connected AI after setup.
No other account needed.
Connect ahel once, and every AI you use reads what you have installed.
Then ask your AI: use the python-ai skill
What this skill tells your AI
The instructions your AI receives, as published by christitustech/titus-ai in .agents/skills/python-ai/SKILL.md and read by ahel’s review.
Workflow
- Gather provider, model, dependency, prompt, tool, and data-flow context.
- Determine privacy, latency, cost, and reliability impact.
- Create rollback with feature flags, fixtures, and provider fallbacks.
- Implement the smallest model, prompt, retrieval, or tool change.
- Validate deterministic code paths and representative AI behavior.
Diagnostics
uv run python --version
uv pip list
uv run pytest
uv run ruff check .
uv run mypy .
env | grep -E 'OPENAI|OPENROUTER|MODEL'
Safety Rules
- Never commit API keys, provider tokens, prompts containing secrets, or private data fixtures.
- Never rely on live model calls for ordinary unit tests.
- Use structured outputs when downstream code depends on response shape.
- Keep tools narrow, deterministic, and logged without secrets.
- Confirm current OpenAI API details from official docs when behavior may have changed.
Validation
- Prompt assembly tests pass.
- Tool schema and parser tests pass.
- Retrieval filters return expected fixtures.
- Optional live-provider smoke tests are gated by environment variables.
- Logs expose model, tool, retry, and error decisions without secrets.
Signals
- GitHub stars
- 125
- Forks
- 21
- Last commit
- Sep 2026
Advanced
- Catalog kind
- skill
- Gateway key
python-ai- Source
- github.com/christitustech/titus-ai