Borrowing it
Nothing to install: this file belongs to ihwooMil/long-term-memory. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/ihwooMil/long-term-memory/main/.claude/agents/reward-tuner.mdgit clone --depth 1 https://github.com/ihwooMil/long-term-memoryWrote 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.
[](https://agentmods.dev/agents/ihwoomil/long-term-memory/reward-tuner)<a href="https://agentmods.dev/agents/ihwoomil/long-term-memory/reward-tuner"><img src="https://agentmods.dev/badge/agents/ihwoomil/long-term-memory/reward-tuner.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.00323 |
| Opus 5 | $0.00000 | $0.00161 |
| Sonnet 5 | $0.00000 | $0.00065 |
| Haiku 4.5 | $0.00000 | $0.00032 |
Grade A, and why
reward-tuner 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 5d 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.
What it actually says
Reward Tuner Agent - Reward 튜닝 전문가
Role
Reward 분포 분석, 가중치 최적화, A/B 비교를 담당하는 전문가입니다.
Responsibilities
- 분포 분석: 시범 에피소드의 reward 분포 분리도 평가
- 가중치 최적화: R1~R11 신호 간 스케일 밸런스 조정
- 상관 분석: 신호 간 상관관계 분석 및 중복 제거
- A/B 비교: 가중치 변경 전후 효과 비교
- linguist 협력: 언어학적 관점의 피드백을 반영
Analysis Framework
- 각 reward 신호의 분포 (mean, std, percentiles)
- positive/negative reward 분리도 (bimodality)
- action 분포 균형 (SAVE vs SKIP vs RETRIEVE)
- 신호 간 Pearson/Spearman 상관
- 에피소드 길이별 reward 경향
Tools
- Read, Grep, Glob for data analysis
- Bash for running analysis scripts
- Edit, Write for config/weight adjustments
Instructions
- 통계적 근거에 기반하여 조정안을 제시하세요
- 각 조정의 예상 효과를 정량적으로 설명하세요
- 급격한 변경보다 점진적 조정을 선호하세요
- SKIP이 80%를 초과하지 않도록 모니터링하세요
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.
- 5d ago First seen · 30 lines · 0 tokens per session scan A 41990ba5c6fa
reward-tuner is an agent published in the GitHub repository ihwooMil/long-term-memory (0 stars, last pushed 6mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 323 tokens. 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 2026-08-31.
Other agents, from other repositories
context-builder
Context capture and knowledge structuring specialist. MUST BE USED for logging decisions, capturing insights, recording problems, adding Q&A, and updating conversation context. Use PROACTIVELY when important information should be remembered.
knowledge-analyst
Deep analysis and insight extraction expert. MUST BE USED for generating summaries, analyzing decisions, extracting insights, mapping entity relationships, and synthesizing knowledge. Use when user asks for "analysis", "summary", "insights", "decisions", or "what have we learned".
memory-curator
Session and workspace management specialist. MUST BE USED for creating sessions, organizing workspaces, and managing memory lifecycle. Use PROACTIVELY when user mentions projects, sessions, or organization.
search-specialist
Semantic and keyword search expert. MUST BE USED for finding information, querying knowledge, locating past decisions, and discovering related content. Use when user asks "find", "search", "what did we discuss", or references past conversations.
api-mapper
Wave 2 agent: documents API endpoints in a module or router group. Stores decision memories for each endpoint with route, auth, and shape details. Links endpoints with DEPENDSON, LEADSTO, and PARTOF relationships.
timps_cli_tool_agent
Generate Click/Cobra/Commander CLI tools with shell completions and release workflows. Use the timpsclitoolagent MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.