菜单
Show Evolver health — evolution memory location & size, recent log, and whether the full engine is installed. Use when the user asks for "evolver status", "evolver health", or "check evolver".
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/EvoMap/evolver-antigravity-plugin --skill evolver-status命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Self-evolution workflow for the agent. Before a substantive task, recall what worked on similar past tasks from evolution memory; after it, record the outcome so future sessions learn from it. Use when the user starts non-trivial work (a feature, a fix, a refactor) or asks the agent to "evolve", "learn from this", or "remember how this went".
Distill a reusable skill/gene from recent run history (optionally from an LLM response file). Use when the user asks to "distill skill", "extract gene", or "generalize lessons".
Run an evolution checkpoint — recall relevant past outcomes, reflect on the current task, and record what was learned. Use when the user requests an evolution checkpoint or asks to "evolve" or "checkpoint".
基于 SOC 职业分类
正在显示 SKILL.md
| name | evolver-status |
| description | Show Evolver health — evolution memory location & size, recent log, and whether the full engine is installed. Use when the user asks for "evolver status", "evolver health", or "check evolver". |
Check the health and readiness of the Evolver persistent self-evolution memory plugin.
When requested by the user, run the following diagnostic checks using shell/terminal commands and report the findings back to the user as a clean checklist:
Evolution memory — check if the local JSONL graph exists and print how many outcomes it holds:
F=~/.evolver/memory/evolution/memory_graph.jsonl
[ -f "$F" ] && echo "memory graph: $F ($(wc -l < "$F" | tr -d ' ') outcomes)" || echo "no local evolution memory yet (it appears after a session ends with changes in a git repo)"
This workspace's id — verify the forge-resistant workspace ID (scoping key):
R=$(git rev-parse --show-toplevel 2>/dev/null); [ -n "$R" ] && { [ -f "$R/.evolver/workspace-id" ] && echo "workspace-id: present" || echo "workspace-id: not yet created (made on first recorded outcome)"; } || echo "not a git repo — memory inactive here"
Recent activity — print the last few lines of the evolution log:
tail -n 5 ~/.evolver/logs/evolution.log 2>/dev/null || echo "no evolution log yet"
Full engine (optional) — check if the global @evomap/evolver CLI is installed:
command -v evolver >/dev/null 2>&1 && evolver --version 2>/dev/null | head -1 || echo "evolver CLI not installed — hooks still work standalone; 'npm i -g @evomap/evolver' unlocks full pipeline features"
Network connection (optional) — check whether a node has been registered locally but not yet claimed on the network:
C=~/.evomap/claim_url
[ -s "$C" ] && echo "not connected yet — claim this node by opening $(cat "$C") while signed in to evomap.ai (that is the only step; no id or secret to enter)" || echo "no pending claim (either not registered yet, or already connected)"
If the Proxy is reachable and evolver_status reports an HTTP 402 / insufficient credits, translate that for the user in plain language: the network features (searching/reusing genes & capsules) need credits — see https://evomap.ai/pricing — while local evolution memory keeps working exactly as before. When reporting connection state, keep it to plain "are you connected?" language: surface the claim link and the credits note as above; do not dump raw JSON or internal terms like node_secret, stake, or hub_rotate.
Finish with a single-line summary statement on the overall readiness of the self-evolution memory.