N memory refresh
Unofficial n-prefixed skills for applying ideas from The Almanack of Naval Ravikant.
npx -y skills add thepraggyverse/naval --skill n-memory-refreshAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Refresh optional Naval memory safely. Use when saved Naval reviews, learnings, decisions, or scorecards need deduplication and freshness review; magic words like "Refresh my saved Naval learnings". Do not use when enable storage or silently rewrite private material.
SKILL.md
2.1 KB, 362 tokens by cl100k_base, as published. Nobody here has run it
Naval Memory Refresh (shared-meta shim)
Pack Context
- Pack config:
owned/shared-meta-skills/pack-configs/naval.yaml - Shared template:
owned/shared-meta-skills/skills/shared-memory-refresh/SKILL.md - Canonical pack:
owned/naval - Pack-local reference:
owned/naval/references/memory/README.md - Direct-install fallback:
references/concept.md - Direct-install fallback:
references/memory.md
Procedure
- Load
memory_refreshandshim_contracts.n-memory-refreshfrom the pack config. - Load only the pack-local references named by the shim and config.
- Follow
shared-memory-refreshfor the invariant workflow and output contract. - Apply the configured pack-specific behavior, fields, and boundary without weakening the shared guardrails.
- Use Naval dispositions keep, update, consolidate, supersede, and recommend-deletion.
Pack-specific behavior preserved
- Resolve the approved memory root and inventory artifact frontmatter by type and tag.
- Group candidates for stale content, contradiction, duplication, weak retrieval, and privacy risk.
- Classify keep, update, consolidate, supersede, or recommend deletion with evidence.
- Present a refresh report and apply no mutation without explicit approval.
Output Contract
- Memory root and counts
- Keep and change classifications
- Contradictions
- Approval-required actions
- Next refresh target
Boundaries
- Fail closed: present a refresh report and mutate only exact entries approved by the user.
- Do not enable storage or silently rewrite private material.
- This skill maintains existing optional memory; it does not enable storage or silently rewrite private material.
Status
Existing Naval meta skill migrated after dual-run coverage proof.
Gives 0 of the 12 instructions most memory context skills give in 362 tokens
Counted across 674 of the 847 authors here whose files we hold, read 2026-08-06
- inform the user when setup is completein 21 of 674, across 6 files
- confirm the draft with the user before writingin 21 of 674, across 6 files
- update the agent skills block in place if it existsin 21 of 674, across 6 files
- present findings to the userin 20 of 674, across 5 files
- write the three docs files from seed templatesin 20 of 674, across 5 files
- ask the user about each decision one at a timein 19 of 674, across 4 files
- edit CLAUDE.md if it existsin 18 of 674, across 3 files
- explore current repo statein 18 of 674, across 3 files
- do not overwrite user edits to surrounding sectionsin 18 of 674, across 3 files
- back up the original file before overwritingin 16 of 674, across 8 files
- keep the memory index under 200 linesin 15 of 674
- Provide actionable steps and verificationin 13 of 674, across 2 files
Said here and by no other author read
- load memory_refresh and shim_contracts from pack config
- load only pack-local references named by the shim
- follow shared-memory-refresh for workflow and output contract
- apply configured pack-specific behavior without weakening guardrails
- use Naval dispositions keep, update, consolidate, supersede, recommend-deletion
- resolve approved memory root and inventory frontmatter by type and tag
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.