TextAI Agents
Claude Code Loops Operator
Turn a coding task into the smallest, safest Claude Code loop primitive that runs until a verifiable stop condition is met.
Prompt CollectionPublic
Agent 与工具调用
工具选择、状态跟踪、副作用确认与失败恢复，适合设计可审计的自动化流程。
By Easy Prompt 编辑部 · 12 prompts
This collection is open to search engine indexing.
TextAI Agents
Auditable Enterprise LLM Agent Harness Architect
Reconstruct prompt-heavy enterprise LLM prototypes into a traceable, auditable, code-owned agent architecture by moving behavior into manifests, schemas, validators, and runtime gates.
TextAI Agents
Agentic Context Engineering Architect
Design self-improving LLM agent context systems that treat context as an evolving playbook, using incremental deltas to defeat brevity bias and context collapse.
TextAI Agents
AgentAtlas Trajectory Eval Architect
Evaluate AI agents by their control decisions and trajectory quality using a six-state taxonomy and failure taxonomy, not just final outcomes.
TextAI Agents
Agent State Hygiene Architect
Designs safe maintenance routines for local AI coding agent state: inspect-before-mutate, archive-don't-delete, and continuity-preserving handoffs.
TextAI Agents
Agent Skill Optimizer Architect
Optimize frozen-LLM skill documents with neural-network training discipline: rollout forward passes, reflection backprop, gradient clipping, and validation-gated acceptance producing deployable best_skill.md.
TextAI Agents
Agent Skill Effectiveness Auditor
Assuming the null hypothesis, run a paired evidence-based audit to decide whether an agent skill actually improves a specific SWE task, weighing context interference and token cost.
TextAI Agents
Agent Environment Engineering Architect
Design the runtime, artifacts, constraints, and interfaces for metric-driven autonomous scientific discovery using off-the-shelf CLI agents.
TextAI Agents
Agent-Native Memory System Architect
Design memory subsystems for long-running agents as first-class data-management systems, covering representation, extraction, retrieval/routing, maintenance, and measurable evaluation.
TextAI Agents
Agent Cost Observability Architect
Design an end-to-end cost-observability and budget-governance system for AI coding agents: cross-provider price normalization, real-time dashboards, budget gates, anomaly detection, and optimization loops.
TextAI Agents
Agent Atlas Trajectory Auditor
Audit LLM agent trajectories using a six-state control-decision taxonomy, separating outcome success, control-decision quality, and trajectory quality.
TextAI Agents
ADHD Parallel Ideation Skill
A parallel divergent ideation skill for coding agents: spawns isolated branches under multiple cognitive frames, scores, clusters, prunes traps, and deepens top survivors.