Install
openclaw skills install skill-upgrade-checkerCheck all installed skills for available upgrades, interpret version changes and risk, and execute upgrades with user confirmation. | 检查所有已安装技能的可用升级,解读版本变更与风险,经用户确认后执行升级
openclaw skills install skill-upgrade-checkerPrimary role: Scan all installed skills for available upgrades, interpret version differences and changelogs, assess upgrade risk, and execute upgrades after explicit user confirmation.
主要职责: 扫描所有已安装技能的可用升级,解读版本差异与变更日志,评估升级风险,经用户明确确认后执行升级。
Before this skill can operate, confirm the following:
clawhub CLI is installed and available in the agent's PATH. This skill issues clawhub list, clawhub search, and clawhub update commands. If the binary is absent, all operations will fail.clawhub login in advance. The clawhub CLI stores credentials in its own default config path (managed by the CLI, not by this skill). No additional environment variables are required by this skill.clawhub list (read-only).clawhub search <skill> (read-only).clawhub update <skill> — only after explicit user confirmation at the Pre-Upgrade Gate (Step 6).
It will not access other system files, credentials, or APIs beyond the clawhub CLI.Run clawhub list and parse the output to build a table of all installed skills with their local version numbers.
Additionally, read each skill's _meta.json (at skills/<skill-slug>/_meta.json) if accessible locally, to get the precise semver from the version field.
Output: a list of installed skills with their local versions.
For each installed skill, run:
clawhub search <skill-slug>
Build a comparison table with columns:
If clawhub search does not return version info for a particular skill (e.g., it is a local-only skill not published to any registry), mark it as "version unknown — skipped".
Apply semver comparison to each skill where an upgrade is available. Classify into:
Edge cases:
Present a summary table sorted by risk level (major first).
For each skill with an available upgrade, gather change information in this priority order:
skills/CHANGELOG.md for entries between the current and latest version.skills/<skill-slug>/CHANGELOG.md if it exists (more specific).description field in local _meta.json with the description returned by clawhub search. Changes in description may hint at capability changes.Summarize changes per skill:
For each upgradeable skill, produce a risk rating using these criteria:
| Dimension | LOW | MEDIUM | HIGH | CRITICAL |
|---|---|---|---|---|
| Version jump | Patch only | Minor | Major | Major with breaking keywords |
| Breaking change signals | None found | "changed", "refactored" | "deprecated" | "breaking", "removed", "incompatible" |
| Dependency impact | No other skill depends on it | 1 skill depends | 2+ skills depend | Core dependency for many skills |
| Version staleness | < 1 month behind | 1–3 months | 3–6 months | > 6 months behind |
| Adoption signals | High install count | Moderate | Low (< 10) | Very new (< 2 weeks) or unknown |
Combine dimensions into an overall rating: LOW / MEDIUM / HIGH / CRITICAL.
Present a prioritized recommendation list, grouped into four tiers:
Include a one-line rationale per skill.
Before executing any upgrade, present a confirmation summary listing:
clawhub update <skill> command to be runThen ask:
"Planned action: upgrade N skill(s). Confirm which to proceed with — respond with skill names, 'all', or 'none'."
Major upgrade extra gate: Even if the user responds with "all", major-version upgrades require individual confirmation per skill. Re-prompt for each major upgrade separately with a clear warning about potential breaking changes.
For each confirmed skill, in sequence:
clawhub update <skill-slug>clawhub list and checking the new versionFinal output:
Upgrade Results Table:
| Skill | Old Version | New Version | Status |
|---|---|---|---|
| ... | ... | ... | Success / Failed |
Post-Upgrade Summary:
| Skill | Local Version | Latest Version | Upgrade Type | Risk Level |
|---|
Grouped by tier (Recommended / Caution / Review / Skip) with one-line rationale.
"Planned action: upgrade <N> skill(s). Source: clawhub. Risks: <summary>. Proceed with: <skill list / all / none>?"
| Skill | Old Version | New Version | Status |