Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add xrgarcia/slashbin-ai-team/plugin install slashbin-harnessWrote 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/xrgarcia/slashbin-ai-team/remember)<a href="https://agentmods.dev/skills/xrgarcia/slashbin-ai-team/remember"><img src="https://agentmods.dev/badge/skills/xrgarcia/slashbin-ai-team/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/skills/xrgarcia/slashbin-ai-team/remember"><img src="https://agentmods.dev/badge/skills/xrgarcia/slashbin-ai-team/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 | $0.00085 | $0.00773 |
| Opus 5.5 | $0.00034 | $0.00309 |
| Sonnet 5.5 | $0.00017 | $0.00155 |
| Haiku 4.5 | $0.00009 | $0.00077 |
Grade A, and why
remember scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.
remember
Answer questions about what happened before, from this bot's own memory.
Why this exists
Without it a bot can only see the last 48 hours of summaries, and on a resumed session not even that — the harness does not re-inject context mid-conversation. So the longer a conversation ran, the less the bot could remember about it. This reaches the whole record instead.
How to do it
Run the gatherer. It searches every store and reports what it found:
node "$CLAUDE_PLUGIN_ROOT/bin/recall.mjs" "<the user's question, or the key terms from it>"
If $CLAUDE_PLUGIN_ROOT is not set, the script sits in bin/recall.mjs beside
this skill.
It prints a Sources searched table followed by cited material. It reads only — it never writes, moves or deletes anything.
Reading the results — this part matters
The table tells you what was actually searched. These are not the same thing:
| result | means |
|---|---|
NO MATCH |
searched, genuinely nothing relevant |
EMPTY |
the store exists but has nothing in it yet |
MISSING / UNAVAILABLE |
never searched — the path or variable is absent |
UNREADABLE |
exists, could not be read |
PARTIAL |
matches found, but some are themselves truncated |
Never present a MISSING source as "I checked and found nothing." That is the
exact failure this replaces: the previous per-repo version read three paths that
had been dead for months, reported that it had checked them, and answered
confidently from the fraction it happened to find.
If the answer would hinge on a source that came back MISSING or UNREADABLE,
say so.
Answering
- Lead with the answer, then the evidence. Not a tour of the search.
- Cite every claim — the summary file and date, or the buffer line. The user should be able to go and look.
- Quote the buffer for anything verbatim. Summaries are compressions and they extrapolate: they invent specificity ("a ~50-line script", "5 test cases") and frame trailing-off threads as decisions. If the question is "what did you say" or "what did I ask", the buffer is the source of truth and the summary is not.
- A
PARTIALbuffer line is truncated. Say the message was cut off rather than reconstructing what it probably said. - If nothing matched, say nothing matched. Do not fill the gap from general knowledge or from the current conversation — a confident wrong recollection is worse than "I don't have that."
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 · 73 lines · 85 tokens per session scan A 60bec5bb324f
remember is a skill published in the GitHub repository xrgarcia/slashbin-ai-team (13 stars, last pushed 6d ago), licensed MIT. It adds 85 tokens to every session and 773 once invoked, about $0.0003 per session on Opus 5.5. 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
memory
Search past conversations in the agent's history log.
learn
Manually record an insight, correction, preference, or rule that the agent should remember for future tasks.
memory-kit
Create a workspace-local .cowork/ memory kit (rules, identity, long-term notes, daily logs, heartbeat templates).
session-logs
Query prior conversations/tasks via the taskhistory tool (replaces filesystem JSONL session logs).
notes-keeper
Read, search, and write notes in the opendray file-system vault. Use this whenever the user mentions "notes", "daily log", "project doc", "TODO list", or wants to persist context that should survive across sessions.
self-improvement
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Claude ('No, that's wrong...', 'Actually...'), (3) User requests a capability that doesn't exist, (4) An external API or tool fails, (5) Claude realizes its knowledge…