Agent Skills
› 410979729/scope-recall-hermes
› scope-recall-setup
scope-recall-setup
GitHub自动执行 Scope Recall 的安装、升级及旧版本记忆迁移。智能识别首次安装或遗留迁移场景,处理环境检查、路径解析、备份及重启命令,全程无需用户干预,确保数据兼容与模型配置保留。
触发场景
用户请求安装或设置 Scope Recall
用户请求更新 Scope Recall
用户请求迁移旧的 Scope Recall 记忆数据
安装
npx skills add 410979729/scope-recall-hermes --skill scope-recall-setup -g -y
SKILL.md
Frontmatter
{
"name": "scope-recall-setup",
"description": "Install or upgrade Scope Recall when the user asks the agent to set it up, update it, or migrate old Scope Recall memories. Automatically distinguish first-time installation from legacy migration."
}
The user asks for the result; perform the technical work as their agent.
- Find the actual host's instance home and Python/plugin environment. Inspect
with
scope-recall setup --host <hermes-or-codex> --home <instance-home>. Check active host configuration first; pass--database <actual-db>for a custom path. Absence at default paths alone must not hide an old database. If the old package lacks this entry point, inspect the new release in an isolated helper environment before touching the running installation. - Follow its route.
fresh_installgoes straight to normal installation and host checks.current_upgradepreserves the current database format.legacy_migrationuses the bundled workflow printed byscope-recall setup --workflow. - Resolve paths, backups, exact audience mapping and restart commands yourself. Do not ask the user to run commands or review individual memories. Do not replace uncertainty about identity/permissions with an owner-private catchall.
- Preserve the user's model choices. Report successful live use separately from package installation, database conversion and pending embedding indexing.
Use the existing migration job and worker rather than generating a second conversion script or repeatedly re-extracting old journals. Preserve failed evidence and stop retrying an unchanged failure.
版本历史
- 97244d5 当前 2026-09-22 05:40


