AgentStack
SKILL verified Apache-2.0 Self-run

Nap

skill-npow-claude-skills-nap · by npow

Context compactor — prunes old sagaflow runs, archives stale memory files, compacts MEMORY.md indexes, compresses aimee journals. Like Squad's `squad nap` but for our Claude Code setup.

No reviews yet
0 installs
14 views
0.0% view→install

Install

$ agentstack add skill-npow-claude-skills-nap

✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

Are you the author of Nap? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

nap

Compact Claude Code context state to free disk and reduce token load.

Usage

When invoked, run the nap script:

python3 ~/code/npow-dotfiles/scripts/nap.py [flags]

Flags from user args

  • If the user says --dry-run or "dry run" or "what would it do": add --dry-run
  • If the user says --deep or "deep" or "aggressive": add --deep
  • If no flags specified: run with --dry-run first, show the report, then run for real

What it does

  1. PRUNE — deletes sagaflow run dirs older than 7 days (3 with --deep)
  2. COMPACT — removes dead pointers from MEMORY.md index files
  3. ARCHIVE — moves memory files not touched in 30 days (14 with --deep) to archive/ subdirs
  4. COMPRESS — archives old aimee journal entries (>7 days or >50KB; >3 days or >20KB with --deep)

Workflow

  1. Run with --dry-run first unless the user explicitly said to just do it
  2. Show the report output to the user
  3. If dry-run, ask once: "Run it for real?" — then execute without --dry-run
  4. Show the final report

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet — be the first.

Versions

  • v0.1.0 Imported from the upstream source.