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/kyopark2014/agent-plugins/memory-managementnpx skills add kyopark2014/agent-plugins --skill memory-managementgit clone --depth 1 https://github.com/kyopark2014/agent-pluginsWhat 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 | $0.00054 | $0.02199 |
| Opus 5 | $0.00027 | $0.01099 |
| Sonnet 5 | $0.00011 | $0.00440 |
| Haiku 4.5 | $0.00005 | $0.00220 |
Grade A, and why
memory-management 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.
This is a copy
97% identical to memory-management — 1 line 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 — 323 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Memory Management
Memory makes Claude your workplace collaborator - someone who speaks your internal language.
The Goal
Transform shorthand into understanding:
User: "ask todd to do the PSR for oracle"
↓ Claude decodes
"Ask Todd Martinez (Finance lead) to prepare the Pipeline Status Report
for the Oracle Systems deal ($2.3M, closing Q2)"
Without memory, that request is meaningless. With memory, Claude knows:
- todd → Todd Martinez, Finance lead, prefers Slack
- PSR → Pipeline Status Report (weekly sales doc)
- oracle → Oracle Systems deal, not the company
Architecture
CLAUDE.md ← Hot cache (~30 people, common terms)
memory/
glossary.md ← Full decoder ring (everything)
people/ ← Complete profiles
projects/ ← Project details
context/ ← Company, teams, tools
CLAUDE.md (Hot Cache):
- Top ~30 people you interact with most
- ~30 most common acronyms/terms
- Active projects (5-15)
- Your preferences
- Goal: Cover 90% of daily decoding needs
memory/glossary.md (Full Glossary):
- Complete decoder ring - everyone, every term
- Searched when something isn't in CLAUDE.md
- Can grow indefinitely
memory/people/, projects/, context/:
- Rich detail when needed for execution
- Full profiles, history, context
Lookup Flow
User: "ask todd about the PSR for phoenix"
1. Check CLAUDE.md (hot cache)
→ Todd? ✓ Todd Martinez, Finance
→ PSR? ✓ Pipeline Status Report
→ Phoenix? ✓ DB migration project
2. If not found → search memory/glossary.md
→ Full glossary has everyone/everything
3. If still not found → ask user
→ "What does X mean? I'll remember it."
This tiered approach keeps CLAUDE.md lean (~100 lines) while supporting unlimited scale in memory/.
File Locations
- Working memory:
CLAUDE.mdin current working directory - Deep memory:
memory/subdirectory
Working Memory Format (CLAUDE.md)
Use tables for compactness. Target ~50-80 lines total.
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 · 323 lines · 54 tokens per session scan A 165c4083833a
memory-management is a skill published in the GitHub repository kyopark2014/agent-plugins (4 stars, last pushed 25d ago), licensed MIT. It adds 54 tokens to every session and 2,199 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. It is 97% identical to memory-management, differing in 1 line, and is treated as a copy.
Other skills, from other repositories
ask-zilliz
Zilliz Cloud onboarding and usage assistant. Helps users understand Zilliz Cloud, choose the right plan, estimate costs, write code, debug issues, and adopt new features like Functions, Volumes, and Global Clusters. Use this skill whenever the user asks about Zilliz Cloud — including plan selection, pricing, cost…
collection
Use when the user wants to create, list, describe, drop, rename, load, release, or manage collections and collection aliases in Milvus.
diagnose
Use when the user reports that a Zilliz Cloud cluster or Milvus collection is unhealthy, slow, stuck, returning errors, hitting quotas, or otherwise misbehaving — or when they ask "what's wrong with...", "why is ... slow", "diagnose ...", "troubleshoot ...".
cluster
Use when the user wants to create, list, describe, delete, suspend, resume, or modify Zilliz Cloud clusters.
external-collection
Use when the user wants to trigger, describe, or list refresh jobs for an external collection (a collection backed by an external data source such as Vector Lake). Note this is the data-plane refresh workflow, not collection CRUD -- for create/drop/load see the collection skill.
on-demand-cluster
Use when the user wants to create, list, describe, or delete an on-demand (Vector Lake / VectorLake) cluster. On-demand clusters auto-suspend after an idle TTL and are intended for ad-hoc query workloads against a Vector Lake.