Install any skill in seconds. Free to start, no credit card required.
Get Started Free →zzafergok/memory-management
Two-tier memory system that makes Claude a true workplace collaborator. Decodes shorthand, acronyms, nicknames, and internal language so Claude understands requests like a colleague would. CLAUDE.md for working memory, memory/ directory for the full knowledge base.
Passed 21 of 22 test cases with the skill, and 7 of 22 without it.
and 18 more in the full report
14 fixed7 passed either way1 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
This benchmark measured an earlier version of the skill.
One of 14 test cases the skill fixed. Same model, same prompt, with and without the skill.
Can you look up what 'PSR' stands for in our team files and tell me who currently handles that report?
| Checked | Without skill | With skill |
|---|---|---|
| Identifies PSR as Pipeline Status Report | Missed | Passed |
| Identifies Todd Martinez as the owner handling the report | Missed | Passed |
The instructions your model reads when the skill loads.
Keep durable context small, auditable, and useful across sessions without turning transient conversation into permanent truth.
Store information that is both reusable and costly to rediscover:
Do not store secrets, credentials, unnecessary personal data, speculation, temporary task state, or information the user did not authorize for persistence.
Use a small always-visible index only for the highest-frequency facts. Place detailed domain knowledge in focused files that can be searched or loaded when relevant. Avoid duplicating the same fact across layers; the index should point to the authoritative detail.
The host agent may use AGENTS.md, a memory directory, project documentation, or another supported mechanism. Follow the environment's documented discovery rules rather than assuming a vendor-specific filename.
Each durable entry should make its status clear:
Distinguish confirmed facts from hypotheses and historical notes.
test cases passed, vs 7 / 22 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.