360 Secure SkillHub
360 安全技能中心

RecallBase

核心功能是从本地存储的AI对话历史中检索有用片段,以自然语言回答问题。需搭配本地rb CLI使用,支持当日对话检索、特定主题搜索等方式。典型场景包括当日工作复盘、恢复中断的AI/编码会话、追溯决策根源。输出基于检索内容,明确标注信息缺失,全程本地运行无需网络登录。

darinrowe
darinrowe
热度 14.45
最近更新 2026-08-20 20:41:40
分类:效率依赖:代码执行 / MCP 服务 / 内部调用大模型应用场景:摘要总结 / 任务管理

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add recallbase
Metadata
name
recallbase
description
Recover context from local AI conversation history with RecallBase. Use for same-day work summaries, resuming prior AI or coding sessions, or tracing a past decision to its source.
license
MIT
metadata
{"author":"Darin Rowe"}

RecallBase

Recover the smallest useful slice of local AI history, then answer the user's question in natural language.

This workflow requires the local rb CLI on PATH and a RecallBase store populated from supported local sources. Core retrieval runs locally without login or network access.

Retrieve

  1. Choose the narrowest entry point:
    • Same-day continuity: rb today --json
    • A known topic, error, file, branch, command, or decision: rb search "<specific query>" --json
    • Suspected coverage gaps: rb sources --json
  2. Inspect the JSON envelope. Continue only from ok: true; for ok: false, use error.code, message, and hint to explain or recover.
  3. Open only the strongest candidate conversations with rb open <conversation-id> --json. Search results also provide the stable reference recallbase:conversation/<id>.

Retrieval is complete when the evidence answers the question or the available source coverage proves the relevant history is absent or incomplete.

Use an already-configured local RecallBase MCP instead of shell commands when it is available. Read references/mcp.md for MCP routing. The public RecallBase Docs MCP contains product documentation; personal history tools come only from the local rb mcp server.

Synthesize

  • Answer the request directly. For a daily recap, group work into themes and name concrete tasks, outcomes, decisions, tests, merged changes, and visible next steps.
  • Treat conversation IDs and rb open ... commands as optional follow-up references, not the answer.
  • Summarize the relevant messages instead of reproducing full transcripts or raw JSON.
  • State material coverage gaps when sources are absent, partial, failed, or stale. When no evidence supports a claim, say so.

The response is complete when every material claim is grounded in retrieved history, uncertainty is explicit, and the user can continue without reading command output.

Conditional references

  • Read references/results.md when interpreting unfamiliar fields or building against the JSON contract.
  • Read references/troubleshooting.md when rb is unavailable, retrieval is empty or incomplete, imports fail, a full re-import is required, or browser capture setup is unhealthy.