rl-policy-optimization

rl-policy-optimization is a skill for Claude Code, Codex from aiming-lab/AutoResearchClaw. It costs 28 tokens per session (351 once invoked), scanned A, original, MIT.

A guide to optimizing policies, which are the decision rules used by reinforcement-learning agents in an environment.

In plain words
What is it for?
Use it when working with PPO, SAC, DQN, or other listed reinforcement-learning methods, including reward design, training logs, evaluations, and parameter sweeps.
Why use it?
It helps choose algorithms and evaluation methods suited to the action type while accounting for unstable rewards, seeds, and hyperparameters.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Use it when working with PPO, SAC, DQN, or other listed reinforcement-learning methods, including reward design, training logs, evaluations, and parameter sweeps.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/aiming-lab/autoresearchclaw/rl-policy-optimization
About the project

AutoResearchClaw is a system that turns a research idea into a scientific paper through autonomous and collaborative AI research workflows. It is for researchers who want agents to investigate questions, run experiments, and produce papers, with optional human guidance. Catalogue skills and agents provide parts of its research workflow.

aiming-lab/AutoResearchClaw · 14,361 stars · on GitHub

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add aiming-lab/AutoResearchClaw --skill rl-policy-optimization
Clone the repo
git clone --depth 1 https://github.com/aiming-lab/AutoResearchClaw

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for rl-policy-optimization

README.md
[![agentmods](https://agentmods.dev/badge/skills/aiming-lab/autoresearchclaw/rl-policy-optimization.svg)](https://agentmods.dev/skills/aiming-lab/autoresearchclaw/rl-policy-optimization)
Your own site
<a href="https://agentmods.dev/skills/aiming-lab/autoresearchclaw/rl-policy-optimization"><img src="https://agentmods.dev/badge/skills/aiming-lab/autoresearchclaw/rl-policy-optimization.svg" alt="Measured on agentmods" height="20"></a>
Per session 28 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 351 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00028 $0.00351
Opus 5 $0.00014 $0.00176
Sonnet 5 $0.00006 $0.00070
Haiku 4.5 $0.00003 $0.00035

Measured 8d ago against content hash 2292f3b42fa2, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

rl-policy-optimization scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

researchclaw/skills/builtin/domain/rl-policy-optimization/SKILL.md · 38 lines

What it actually says

RL Policy Optimization Best Practice

Algorithm selection:

  • Discrete actions: PPO, DQN, A2C
  • Continuous actions: SAC, TD3, PPO
  • Multi-agent: MAPPO, QMIX
  • Offline: CQL, IQL, Decision Transformer

Training recipe:

  • PPO: clip=0.2, lr=3e-4, gamma=0.99, GAE lambda=0.95
  • SAC: lr=3e-4, tau=0.005, auto-tune alpha
  • Use vectorized environments (e.g., gymnasium.vector)
  • Normalize observations and rewards
  • Log episode return, episode length, value loss, policy entropy

Evaluation:

  • Report mean +/- std over 10+ evaluation episodes
  • Use deterministic policy for evaluation
  • Compare against random policy and simple baselines
  • Report sample efficiency (return vs. env steps)

Common pitfalls:

  • Reward shaping can introduce bias
  • Seed sensitivity is HIGH — use 5+ seeds
  • Hyperparameter sensitivity — do a small sweep
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 8d ago First seen · 38 lines · 28 tokens per session scan A 2292f3b42fa2

Subscribe to this mod's changes

rl-policy-optimization is a skill published in the GitHub repository aiming-lab/AutoResearchClaw (14,361 stars, last pushed 20d ago), licensed MIT. It adds 28 tokens to every session and 351 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories