Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add TashanGKD/tashan-cursor-skills --skill import-ai-memorygit clone --depth 1 https://github.com/TashanGKD/tashan-cursor-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/tashangkd/tashan-cursor-skills/import-ai-memory)<a href="https://agentmods.dev/skills/tashangkd/tashan-cursor-skills/import-ai-memory"><img src="https://agentmods.dev/badge/skills/tashangkd/tashan-cursor-skills/import-ai-memory/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/tashangkd/tashan-cursor-skills/import-ai-memory"><img src="https://agentmods.dev/badge/skills/tashangkd/tashan-cursor-skills/import-ai-memory.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00129 | $0.03287 |
| Opus 5 | $0.00064 | $0.01643 |
| Sonnet 5 | $0.00026 | $0.00657 |
| Haiku 4.5 | $0.00013 | $0.00329 |
Grade A, and why
import-ai-memory scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
This is a copy
80% identical to smart-search — 337 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 288 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI 记忆导入与整合
触发时机
用户将 AI(如 ChatGPT、Claude、Gemini)根据提示词返回的回复粘贴过来,请求整合进科研数字分身。若用户使用多个 AI 工具,会依次粘贴多段回复。
重要:不要在流程开始时询问用户姓名或标识。直接从步骤一解析 AI 回复开始。
步骤一:解析 AI 回复内容
读取用户粘贴的 AI 回复(可能包含多段,来自不同 AI),按以下方式解析:
- 若用户粘贴了多段回复,按粘贴顺序依次解析,合并同字段信息(后粘贴的补充或覆盖先前的,冲突时标记)
- 识别回复中涉及的模块(A 基础身份 / B 能力 / C 当前需求 / D 认知风格 / E 动机人格)
- 对每条信息,记录 AI 标注的可信度标签及依据(若有):
- ✅ 有据可查(含 AI 提供的证据/原话摘录)
- ⚠️ 印象模糊
- ❌ 记忆不足
- 读取当前用户的画像,对比每条 AI 信息与现有数据:
- 画像为空白:该信息为新增内容
- 画像已有数据:需标注冲突
- 内容一致:可直接采纳
【机制一】❌ 记忆不足条目不丢弃:将所有被标记为 ❌、且对应画像字段仍为空的条目,记录进内部「待补充字段列表」,留待步骤五·5.2 发起追问。不得在此处跳过或忽略。
步骤二:分类处理
有据可查且无冲突的条目:内部整合后直接写入,不向用户展示,写入时标注置信度(见步骤四规则)。
仅对以下类型向用户发起确认(优先选择题,必要时填空或问答并给回答样例):
- ⚠️ 印象模糊 的条目
- 与画像已有数据存在冲突 的条目
- C 模块(当前需求) 的所有条目(无论可信度,因属高度个人化信息)
步骤三:逐条确认(选择题优先,必要时填空/问答+样例)
对需要确认的条目,优先使用选择题;若选择题无法覆盖,再用填空或开放问答,并给出回答样例。
选择题格式(优先)
[字段名称]
AI 的回答:「[原文摘录]」(可信度:⚠️ 印象模糊 / 与画像冲突 / C模块)
[若为冲突:你之前填写的:「[画像中的原始数据]」]
→ 请选择:
A. [选项一,如:基本符合,可以写入]
B. [选项二,如:不太准确,我来补充]
C. [选项三,如:跳过这条]
冲突情况的特殊处理(选择题)
⚠️ 此条信息与你之前填写的数据存在出入
AI 的回答:「[原文]」
你之前填写的:「[画像中的原始数据]」
→ 请问你希望如何处理?
A. 保留我之前填写的内容(AI 记忆有误)
B. 以 AI 的回答为准(AI 说得更准确)
C. 两者都有参考价值,帮我合并(请说明如何合并,例:取 AI 的机构名 + 我之前的领域描述)
D. 暂时跳过,我需要想一想
填空或开放问答(必要时,须附回答样例)
当选择题无法充分表达时,使用填空或问答,并给出回答样例:
[字段名称]
AI 的回答比较模糊:「[AI 原文]」
→ 请补充或修正(若无需修改可回复「保持原样」):
样例:我目前在 XX 大学读博,导师做计算神经科学方向。
💡 「当前需求」是画像中最个人化的维度,AI 的推断仅供参考。
[字段名称] AI 的回答:「[原文]」
→ 这条描述是否真实反映了你现在的状态?
A. 是,可以写入
B. 否,我来补充正确信息(例:我最近主要精力在写毕业论文,卡在实验数据整理)
C. 跳过这条
步骤四:写入画像
用户完成所有需确认条目的选择/补充后,执行写入:
- 有据可查且无冲突的条目:已内部整合,直接写入
- 用户确认通过的模糊/冲突/C 模块条目:写入
- 在调用 write_profile 之前:若尚未确定用户姓名或标识,此时单独询问:「请提供您的姓名或标识,用于命名/保存科研数字分身」。用户提供后再执行写入。
- 使用
write_profile工具更新会话中的画像(Web 模式)或StrReplace更新profiles/[姓名].md(Cursor 模式)
【机制三】来源标注规则(必须严格区分,不得混用):
写入路径 字段标注 ✅ 有据可查,静默写入(用户未审核) (来源:AI记忆,置信度高,待审核)⚠️ 模糊/冲突,用户确认通过 (来源:AI记忆,已用户确认)C模块,用户确认通过 (来源:用户确认)步骤五追问中用户填写 (来源:用户补充)禁止将 ✅ 静默写入的条目标注为"已用户确认"。
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 288 lines · 129 tokens per session scan A 8a7a2673b931
import-ai-memory is a skill published in the GitHub repository TashanGKD/tashan-cursor-skills (20 stars, last pushed 5mo ago), licensed MIT. It adds 129 tokens to every session and 3,287 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. It is 80% identical to smart-search, differing in 337 lines, and is treated as a copy.
Other skills, from other repositories
cognitive-daily-briefing
A read-only daily summary of recent changes, unfinished items, notes, and the current state of a personal knowledge system.
notion
Read, create, and update Notion pages and databases via the Notion REST API.
hermespace
Full Hermespace pocket workbench for Hermes Agent (v0.18+): concept, dual decode, workbench I/O, fabric skills+MEMORY, FOA desk, Pulse, viewport app-shell, Tailscale serve, pocket boundary, autonomy grid, Desktop plugin, ops boot/doctor/smoke. Use for hermespace, pocket, workbench, FOA, pulse, viewport, access…
cognitive-consistency-check
A full validation of a structured knowledge system against ten consistency rules, including existing references, declared relationships, document status, and non-contradictory content.
cognitive-extract-principle
A Chinese-language skill for finding recurring patterns across separate notes and turning them into candidate underlying principles.
cognitive-update-knowledge
A controlled workflow for updating a core knowledge document, including impact analysis, backups, consistency checks, and related record updates.