Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/tstockham96/claw-kitnpx agentmods add commands/tstockham96/claw-kit/rememberWrote 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/commands/tstockham96/claw-kit/remember)<a href="https://agentmods.dev/commands/tstockham96/claw-kit/remember"><img src="https://agentmods.dev/badge/commands/tstockham96/claw-kit/remember/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/commands/tstockham96/claw-kit/remember"><img src="https://agentmods.dev/badge/commands/tstockham96/claw-kit/remember.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 | $0.00000 | $0.01105 |
| Opus 5.5 | $0.00000 | $0.00442 |
| Sonnet 5.5 | $0.00000 | $0.00221 |
| Haiku 4.5 | $0.00000 | $0.00111 |
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 1mo 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 — 135 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/remember
Store something in the appropriate memory file. This command is the smart router for the memory system.
Input
Read the user's input from $ARGUMENTS. This is the thing they want to remember.
Instructions
Step 1: Parse and Classify
Analyze the input and classify it into exactly one of these categories:
| Category | Target File | Trigger Examples |
|---|---|---|
| Fact | memory/long-term.md |
General knowledge, reference info, account details, important dates |
| Preference | memory/preferences.md |
Likes, dislikes, style choices, tool preferences, workflow habits |
| Person | memory/people/[name].md |
Info about a specific person -- who they are, relationship, contact details |
| Project | memory/projects/[name].md |
Project status, context, goals, tech stack, deadlines |
| Decision | memory/decisions/YYYY-MM-DD-slug.md |
A significant decision with rationale -- "we decided to...", "I chose..." |
| Lesson | memory/learnings.md |
Mistakes made, corrections received, patterns discovered |
If the classification is ambiguous, present the top two options and ask the user which fits best. Do not guess.
Step 2: Check for Duplicates
Before storing, search memory to see if this information already exists:
cd kit/search && npx tsx src/cli.ts search "key terms from the input" --limit 5 --memory-path ../memory
If a close match is found, tell the user and ask whether to update the existing entry or add a new one.
Step 3: Route and Store
Based on the classification:
Fact
- Read
memory/long-term.md - Append the fact under the appropriate section (
## Factsor## Reference) - Include today's date in parentheses after the entry:
(YYYY-MM-DD)
Preference
- Read
memory/preferences.md - Append under the most fitting section (
## Communication,## Technical, or## Work Style) - If no section fits, create a new
## Othersection - Include today's date in parentheses
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.
- 1mo ago First seen · 135 lines · 0 tokens per session scan A f10df344966b
remember is a command published in the GitHub repository tstockham96/claw-kit (5 stars, last pushed 8mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,105 tokens. 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 2026-08-31.
Other commands, from other repositories
tour
Tour this project's memory — what's captured, where it lives, how to recall it (core-memory-kit).
help
Show how to use Penfield memory.
learn
Extract patterns and learnings from current session.
export-closedloop-learnings
Exports pending ClosedLoop learnings to global location with deduplication.
log
View memory access log (reads, writes, deletes).
setup
Configure mnemonic for proactive memory operations.