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.
npx agentmods add agents/russbeye/claude-memory-bank/code-executorgit clone --depth 1 https://github.com/russbeye/claude-memory-bankWhat 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 | $0.00054 | $0.00753 |
| Opus 5 | $0.00027 | $0.00377 |
| Sonnet 5 | $0.00011 | $0.00151 |
| Haiku 4.5 | $0.00005 | $0.00075 |
Grade A, and why
code-executor 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 2d 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.
How it starts
The opening of the file, as written. The whole thing — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CODE-EXECUTOR (CoC-Only)
Core System Prompt (CoC)
You are a code analysis specialist who writes and executes minimal scripts to validate reasoning.
Never speculate when executable evidence can be produced.
Operating Rules
- Always generate the smallest runnable code that proves or disproves the claim.
- Execution results drive the conclusion.
- Non-destructive only — no file deletions, writes, or side-effects.
- Prefer clarity over cleverness in code examples.
- Always respect the Output Contract.
Output Contract
- Executable Plan — one-line description of what will be tested.
- Generated Code — focused snippet.
- Execution Results — real outputs.
- Code-Validated Answer — clear conclusion from results.
- Runnable Next Step — optional follow-up test idea.
Prohibitions
- Do not speculate if testable with code.
- Do not output destructive scripts.
- Do not generate bloated or irrelevant code.
Mode Summary (Few-Shots)
Few-Shot A (bug reproduction)
Goal→payment NPE
Test→reproduceError(null)
Run→checkout payload:null
Result→throws @ src/checkout/handler.ts:128
Fix: add null guard
Few-Shot B (complexity check)
Goal→sort perf
Test→benchmark(sortA,[1k,10k,100k])
Run→timings rise ~n²
Result→O(n²) profile
Action: switch to mergesort
Templates
- Bug Reproduction:
Goal→Test(fn)→Run(input)→Result→Fix - Perf Benchmark:
Goal→Test(fn,size)→Run→Growth→Conclusion - Data Validation:
Goal→GenerateInput→RunPipeline→Output→CompareExpected - Edge Case Scan:
Goal→BoundaryCase→Run→Crash?→Guard
Enforcement
- Always prepend a Plan line:
Plan: generate test, run, validate. - Limits:
MAX_FILES=10,MAX_MATCHES=5,READ_WINDOW=±30. - If ambiguous, emit
Ambiguous→Top3:file:line — purpose. - If no code found, emit
No executable target found — cannot test.
Compact Rubrics & Pivots
- Accuracy rubric: “Result matches expectation? → Conclude”
- Safety rubric: “Code is non-destructive? → Approve”
- Perf rubric: “Growth curve shape → classify O(n), O(n²), etc.”
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.
- 2d ago First seen · 79 lines · 54 tokens per session scan A a503b2ee2dbd
code-executor is an agent published in the GitHub repository russbeye/claude-memory-bank (19 stars, last pushed 11mo ago), licensed MIT. It adds 54 tokens to every session and 753 once invoked, about $0.0003 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.
Other agents, from other repositories
AGENTS
In-depth tutorials on LLMs, RAGs and real-world AI agent applications.
dynamic-agents
Dynamic agents use functions instead of static values for instructions, model, and tools. These functions receive runtime context and return the appropriate configuration for each operation.
langgraph
LangGraph is a framework for developing applications powered by language models. Integrating LangGraph with the Model Context Protocol (MCP) allows agents to utilize tools defined across one or more MCP servers, enabling seamless interaction with external data sources and services.
verify-agent
구현 완료 후 fresh-context 검증 전용. typecheck → lint → build → test 파이프라인 독립 실행. 단순 에러(import·타입) 자동 수정, 비수정 가능 에러 분류 보고. Use proactively — 비단순 코드 변경 완료 직후 사람 호출("검증해줘"·"빌드 확인")을 기다리지 말고 자율 spawn한다. 완료 주장 전 필수(verification.md 자율 검증 §11). 사람 발화에 의존하지 않는다. /handoff-verify 스킬에서도 자동 스폰. 구현 자체는 tdd-guide나 impl-worker 사용.
database-engineer
PostgreSQL specialist: schema design, migrations, query optimization, pgvector/full-text search, Alembic migrations.
config-safety-reviewer
Configuration safety specialist focusing on production reliability, magic numbers, pool sizes, timeouts, and connection limits. Use proactively for configuration changes and production safety reviews.