cy0307/awesome-loop-engineering
Awesome Loop Engineering Dataset A structured dataset of 1022 papers, official docs, tools, benchmarks, patterns, critiques, and implementation guides for recurring AI-agent systems. Resource Atlas · GitHub field guide · Resource selection · Report a correction Dataset Summary Each row connects an original source to its contribution, novelty, impact, publication details, lifecycle stages, audience, evidence type, link status, and… See the full description on the dataset page: https://huggingface.co/datasets/cy0307/awesome-loop-engineering.
34.5k
1# Security Policy2 3Awesome Loop Engineering ships documentation, data, examples, and scripts rather than a production service or hosted runtime. Security concerns can still appear in executable starters, schemas, links, or contributed material.4 5## Please Report6 7- A linked resource that appears malicious or impersonates another project.8- A script or example that could cause unsafe actions if copied directly.9- Accidentally committed credentials, tokens, private URLs, or sensitive data.10- Guidance that could encourage unsafe production autonomy without approvals or escalation.11 12## How To Report13 14Open a private security advisory if GitHub offers that option for this repository. If not, open an issue with sensitive details removed and ask for a maintainer response.15 16Do not post secrets, private customer data, exploit details, or internal URLs in public issues or pull requests.17 18## Security Review Standard19 20Loop Engineering examples should be conservative:21 22- sensitive actions require human approval;23- production actions should be read-only by default unless explicitly scoped;24- credentials and secrets must never be included in examples;25- loops should have retry budgets, exit conditions, and escalation paths;26- verification should rely on concrete evidence, not only model self-assessment.27