Instructions file CodexOpenCode
Instructions for simple-agent-lab/AutoTrainess, covering agents.md, objective, hard constraints, experiment guidance and stage rules.
21 29d ago A 1,265 tokens
original MIT
data
02Skill Claude CodeCodex
Use when preparing training data.
21 29d ago A 8 tokens
original MIT
eval
03Skill Claude CodeCodex
Use when evaluating a model on the current benchmark(s).
21 29d ago A 13 tokens
original MIT
Skill Claude CodeCodex
Use when defining the goal and action plan for the next experiment iteration.
21 29d ago A 17 tokens
original MIT
log
05Skill Claude CodeCodex
Use when appending an experiment log entry after a completed iteration.
21 29d ago A 15 tokens
original MIT
train
06Skill Claude CodeCodex
Use when running benchmark-oriented training with LlamaFactory, including SFT and RL.
21 29d ago A 19 tokens
original MIT