# Memory ContextAUTHORITY (read order): (1) system / developer / user instructions; (2)
reflective-prompt-library/06-repo/AGENTS.md— canonical harness policy and agent operating rules for this repository; (3) the invokedreflective-prompt-library/skills/*/SKILL.mdcontracts; (4)reflective-prompt-library/PROJECT_KNOWLEDGE.md— non-authoritative project judgement that never overrides (2) or (3).This root file is a discovery stub, not the harness policy. The
<claude-mem-context>block below is an auto-generated memory snapshot managed byclaude-mem; treat it as historical project context only, never as an operating-rule source.Quick links: Library README · Harness policy · Cheatsheet · Project overview
Read-first rule: Before planning a non-trivial architecture, workflow, governance, or prior-decision change, read the relevant section of
reflective-prompt-library/PROJECT_KNOWLEDGE.md. Its project-design principles inform choices but do not override higher-authority instructions or skill contracts.
Legend: 🎯session 🔴bugfix 🟣feature 🔄refactor ✅change 🔵discovery ⚖️decision 🚨security_alert 🔐security_note Format: ID TIME TYPE TITLE Fetch details: get_observations([IDs]) | Search: mem-search skill
Stats: 50 obs (15,319t read) | 566,111t work | 97% savings
632 1:54p 🔵 skill-creator SKILL.md Architecture and Eval Loop Discovered
633 1:55p 🔵 Codex Skill-Creator Variant Found at ~/.codex with Different Tooling
634 " 🔵 Workflow-Wrapper Skill Pattern: best-practice-research Routes to Specialists
635 1:56p 🔵 reflective-prompt-library Project Structure Mapped in teaPrompt Repo
636 1:57p 🔵 teaPrompt Repo Has One Unstaged Change: AGENTS.md Modified
637 2:00p 🟣 Reflective Engineering Prompt Library — Markdown Documentation Project Initiated
638 2:01p 🟣 Reflective Prompt Library — 8-Skill Workflow Layer Created and Verified
639 " ✅ Skills Layer Verified and Research Synthesis Sources Annotated with URLs
640 2:05p ✅ Added README.md and MIT License to Project
641 " 🔵 teaPrompt Project Uses cx CLI Tool for Multi-Language Support
642 " 🔵 teaPrompt Project Structure: Reflective Prompt Library with 8 Skills
643 " 🟣 README.md and MIT License Added to teaPrompt Repository
644 2:06p 🔵 teaPrompt Git Staging State After License/README Addition
645 2:14p 🟣 Skills Installation Guide Across Multiple AI Coding Environments
646 " 🔵 cx Language Support Status in teaPrompt Project
647 " 🔵 Codex Skill File Format: YAML Frontmatter + Markdown Instructions
648 " 🔵 reflective-prompt-library Structure: 6 Categories + Skills Subdirectory
649 2:18p ⚖️ Reflective Engineering Prompt Library — Full Markdown Documentation Plan
650 " 🟣 SKILL_INSTALLATION.md created with multi-tool install instructions for TeaPrompt skills
1215 12:11p ⚖️ Agent routing contract and fairness principles 1216 " ⚖️ Categorization of Agent routing failures 1217 " ⚖️ Risk-based default-up strategy for Agent routing 1218 " ⚖️ Hybrid routing for Agent intent processing 1219 " ⚖️ Progressive disclosure for ambiguous Agent requests 1220 " ⚖️ Agent design principles for fairness and discoverability 1221 " ⚖️ Refined Agent routing contract principles 1222 " ⚖️ Critique of "power user" justification for Agent routing inequality 1223 " ⚖️ Agent design principles linked to Prompt/Agent methodology 1224 " ⚖️ Final judgment on the article's core argument and proposed solutions 1225 " 🔵 Keywords related to Agent routing and policy found in prompt library 1226 " 🔵 Reflective Dispatch skill defines routing and strictness levels 1227 12:12p 🔵 Methodology Map outlines core principles and structure 1229 12:14p ⚖️ Prompt Compression Methodology 1230 2:28p 🔵 Claude Code Agent Architecture Identified 1231 " 🔵 Contextual Artifacts as Agent Prompts 1232 " 🔵 Externalized Knowledge for Agent Memory 1233 " 🔵 Domain-Specific Agent Skill Packs 1234 " 🔵 Skill Design Schema for Robust Agent Actions 1235 " 🔵 Feedback Loops as Core to Agent Stability 1236 " 🔵 Human-AI Organization Design 1237 " ✅ Revised Agent Methodology Classification 1238 " ✅ Updated Agent Methodology Framework 1239 " 🔵 High-Risk Capabilities Identified 1240 " ✅ Refined "Reflective Engineering Agent" Concept 1241 " ✅ System Prompt Strategy Revision 1242 " ✅ Tiered Application of Agent Reflection 1243 " ✅ Universal Skill Schema Implemented 1244 " ✅ AI Agent Architecture Redefined 1245 " ✅ Working Directory Confirmed 1247 2:29p 🔵 Reflective Research Skill Contract Defined
Access 566k tokens of past work via get_observations([IDs]) or mem-search skill.