Panniantong/Agent-Reach A Agent Skill Active
Multi-platform search CLI for 17 sites including Chinese platforms
19 MCP servers and agent skills in the Context Engineering category, ranked by quality score.
Multi-platform search CLI for 17 sites including Chinese platforms
Understand what context is, why it matters, and the anatomy of context in agent systems
Recognize patterns of context failure: lost-in-middle, poisoning, distraction, and clash
Design and evaluate compression strategies for long-running sessions
Apply compaction, masking, and caching strategies
Master orchestrator, peer-to-peer, and hierarchical multi-agent architectures
Design short-term, long-term, and graph-based memory architectures
Build tools that agents can use effectively, including architectural reduction patterns
Build evaluation frameworks for agent systems
Widely used prompt engineering techniques and patterns, including Anthropic best practices and agent persuasion principles.
Preserves the reasoning behind a codebase — decisions, workarounds, rejected alternatives
`mfs-find` / `mfs-ingest` skills that search, grep and read across your code, docs, chat (Slack/Gmail/Jira), databases and object stores as one file-like, searchable namespace; self-hosted with local ONNX embeddings
Token audit, usage tracking, and swipe-to-delete skill pruning.
Graph-based long-term memory skill for AI (LLM) coding agents — faster context, fewer tokens, safer refactors
Persistent memory with hooks, wiki, and daily synthesis for multi-project workflows
Maintains portable, cited agent knowledge bases in plain Markdown