AGENT SKILL

rag-memory

AgentScope retrieval, vector-store, and long-term memory workflows.

License 许可
Apache-2.0
最近更新
2026-09-03
Stars
★ 351
Forks
27

01 前言 Front Matter

name
rag-memory
license
Apache 2.0
metadata
{ "disco-role": "operating" }
description
AgentScope retrieval, vector-store, and long-term memory workflows.
disable-model-invocation
true

02 资源文件 6

03 SKILL.md 正文

rag-memory

Use this sub-skill for document indexing, vector search, agent-attached RAG, filesystem memory, mem0, and ReMe.

Read first

  • references/workflows.md for the end-to-end indexing and memory patterns.
  • references/vector-stores.md for Qdrant, Milvus Lite, MongoDB, and Elasticsearch choices.
  • references/troubleshooting.md for dimension, backend, and memory-mode failures.
  • The bundled scripts in scripts/ for local demos.

Typical triggers

  • "How do I index documents and search them?"
  • "How do I attach RAG to an agent?"
  • "How do filesystem memory, mem0, or ReMe work?"
  • "Which vector store or embedding dimension should I use?"

What belongs here

  • KnowledgeBase, TextParser, ApproxTokenChunker, QdrantStore
  • RAGMiddleware
  • AgenticMemoryMiddleware, Mem0Middleware, ReMeMiddleware
  • vector-store selection and memory-backend selection
  • safe RAG and filesystem-memory demos

What does not belong here

  • provider credentials and model-family selection → provider-connectors
  • agent/tool/permission basics → agent-core
  • service bootstrap, storage, and deployment → service-platform
  • workspace backend configuration → workspace-sandboxes

Use pattern

  1. Decide whether the workflow is pure retrieval or retrieval plus an agent.
  2. Pick the vector store or memory backend that matches the deployment target.
  3. Match the embedding dimension before changing the code.
  4. Use the bundled scripts for the local demos first.
  5. Read the troubleshooting page before touching the backend config if a memory or vector call fails.

Bundled scripts

  • scripts/index_and_search.py — standalone RAG indexing/search demo.
  • scripts/integrate_with_agent.py — attach RAGMiddleware to an agent.
  • scripts/agentic_memory_demo.py — filesystem-backed long-term memory demo.

Cross-links

  • If the embedding model itself is the issue, switch to provider-connectors.
  • If the task is really about workspace file placement or sandboxed backends, switch to workspace-sandboxes.
  • If the task is about service deployment or KB endpoints, switch to service-platform.

04 同仓库更多 Skills

VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-triton-opt

Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-solve

END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-skill

PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-pattern-fusion

Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-orchestrate

Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-feedback

Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).

★ 351