llman 0.0.65

A tool for managing LLM application rules(prompts) ...
Documentation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
# agentdev/

This directory contains developer-facing assets for **agent workflows** and **prompt evaluation**.

Scope:
- Prompt evaluation fixtures (Promptfoo)
- Runner scripts (local + docker)
- Experiment notes and tooling that are **not** part of the shipping `llman` CLI

Non-goals:
- Do not store user-facing runtime assets here.
- Do not store test config fixtures here (those live under `artifacts/testing_config_home/`).
- Do not commit secrets or credentials.

If you are looking for Promptfoo suites, see `agentdev/promptfoo/`.

Docker runners:
- `agentdev/docker/sdd-claude-style-eval/`: reproducible env for Claude Code agentic multi-style eval (supports mirror build args)