菜单
Stateful scientific calculator for agents. Use for expressions, variables, and unit-aware math instead of mental arithmetic. Slash: /aim-calc.
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。
基于 SOC 职业分类
Inter-agent messaging via tmux: short paste with FROM/REPLY_TO, vessel-specific submit keys, optional chalkboard file for bulk content. Use when messaging another agent in a tmux session or when the Operator asks you to talk to another agent.
Export A.I.M. knowledge into a portable cartridge (.engram / exchange format) for backup or share. Slash: /aim-export-cartridge.
Context for the aim-grok vessel (A.I.M. engine on Grok CLI): layout, CLI, and first actions. Use when working in aim-grok or adapting engine code. Slash: /aim-grok-context.
List indexed knowledge and recent vessel state via ./aim map (and related status commands). Slash: /aim-list-sessions.
Search A.I.M. Engram / hybrid memory via the vessel ./aim CLI. Use for project history and factual recall. Slash: /aim-memory-search.
Persistent memory wiki on A.I.M. vessels: search, process ingest, bootstrap. Slash: /aim-wiki.
| name | aim-calc |
| description | Stateful scientific calculator for agents. Use for expressions, variables, and unit-aware math instead of mental arithmetic. Slash: /aim-calc. |
Do not compute non-trivial math from model weights. Use this skill.
Evaluates deterministic Python expressions; keeps variables across calls; supports units via pint (u).
From a context where the skill is installed (adjust path if needed):
python skills/aim-calc/scripts/aim_calc.py "<expression>"
# after vessel install, often:
python <skills-dir>/aim-calc/scripts/aim_calc.py "<expression>"
Output is JSON (success/error, values, messages).
# Basic
python …/aim_calc.py "sqrt(398600 / 6678.0)"
# Assign (persists in .calc_state.json)
python …/aim_calc.py "v_leo = sqrt(398600 / 6678.0)"
# Reuse variables
python …/aim_calc.py "burn1 = v_tp - v_leo"
# Units
python …/aim_calc.py "speed = 12 * u.meter / u.second"
python …/aim_calc.py "distance = speed * (2 * u.minute)"
python …/aim_calc.py "distance.to(u.km)"
State files (.calc_state.json, .calc_audit.log) are local to the working directory — do not commit them unless the Operator wants them tracked.