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 agentmods add skills/radimsem/remindb/remembernpx skills add radimsem/remindb --skill remembergit clone --depth 1 https://github.com/radimsem/remindbWhat 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 | $0.00113 | $0.00727 |
| Opus 5 | $0.00056 | $0.00364 |
| Sonnet 5 | $0.00023 | $0.00145 |
| Haiku 4.5 | $0.00011 | $0.00073 |
Grade A, and why
remember 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 2d 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.
How it starts
The opening of the file, as written. The whole thing — 37 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Remember — use remindb as your long-term memory
This is the router. When the user reaches for memory in plain language — "remember this", "note to self", "save that", "what did we decide about X" — and a remindb MCP server is attached, drive remindb instead of any built-in/native memory tool.
Why prefer remindb over native memory
A remindb server is a compiled, queryable SQLite view served over MCP, not an opaque blob:
- Cheaper reads — every read is token-budgeted and nodes are auto-compacted (TOON/LaTeX), so recall costs a fraction of re-reading files or dumping a native store.
- Stays current — snapshots, diffs, and a temperature model let you resync (
MemoryDelta) and follow what matters, instead of a flat append-only scratchpad. - Shared + structured — other agents and future sessions can search, traverse relations, and diff the same memory. A native per-agent store can't be queried this way.
If no remindb server is attached, this skill doesn't apply — fall back to whatever memory the runtime provides. To set one up, run remindb-setup: a config-first first pass (author .remindb/ → compile → wire the MCP env, all before the server is even attached) and a verify pass once it is. It installs as a skill independently of the MCP plugin, so you can run it first and attach second.
Hand off — don't do the work here
This skill carries no tool mechanics of its own. Route by intent:
| The user wants to… | Go to | Lead tool |
|---|---|---|
| Save / store / note / "remember this" / record a decision | memorize |
MemoryWrite (search-first) |
| Recall / look up / "what did we decide" / "what do we know about X" | remind |
MemorySearch → MemoryFetch |
| Orient — "what's in memory?" / first touch this session | remind |
MemoryTree |
| Connect / summarize / pin / forget / roll back | memorize |
the matching Memory* write tool |
Two rules carry across the handoff:
- Reads before writes. Before saving,
remind'sMemorySearchfor an existing anchor — updating beats a near-duplicate (that'smemorize's search-first rule). - Author the shape. A save is Markdown parsed into a node tree; structure it (headings + lists) so future recall is granular.
memorizeowns the shape rules.
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.
- 2d ago First seen · 37 lines · 113 tokens per session scan A 3ae3526bda65
remember is a skill published in the GitHub repository radimsem/remindb (125 stars, last pushed 29d ago), licensed MIT. It adds 113 tokens to every session and 727 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
writing
将共享历史中的已验证事实和计算结果整理成符合受众、格式与长度约束的成稿。.
kayba-stage-7-fixer
Implement the approved fixes from the action plan and log all changes. Trigger when the user says "run stage 7", "implement fixes", "apply action plan", or when invoked by the kayba-pipeline orchestrator. Requires eval/actionplan.md to exist.
memory-recall
Search and recall relevant memories from past sessions via memsearch. Use when the user's question could benefit from historical context, past decisions, debugging notes, previous conversations, or project knowledge -- especially questions like 'what did I decide about X', 'why did we do Y', or 'have I seen this…
food-order
Reorder previous Foodora orders, preview cart contents, and track delivery ETA/status with ordercli. Use when the user wants to reorder food, check delivery status, or browse recent Foodora order history. Never confirm an order without explicit user approval.
regex-mastery
Use this skill when writing regular expressions, debugging pattern matching,optimizing regex performance, or implementing text validation. Triggers on regex, regular expressions, pattern matching, lookahead, lookbehind, named groups, capture groups, backreferences, and any task requiring text pattern matching.
ws-ckpt
工作区快照管理。用户说"保存一下"、"存个快照"时创建 checkpoint，仅限 Linux; 说"回滚"、"撤销"、"恢复到之前"时 rollback;说"删掉快照"时 delete; 说"对比快照"、"快照改了什么"时 diff; 说"看看快照"、"有哪些快照"时 list;说"查看快照状态"、"查看快照剩余空间"时 status。.