Case 02963

Skill knownasnaffy/prompthound/dataset/case_02963

A fast, offline static risk analysis CLI for AI agent skill files. Detects malicious instructions, steganographic payloads, and dangerous capability chains.

Install

npx -y skills add knownasnaffy/prompthound --skill case_02963

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
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

Convert Elite Longterm Memory data into a structured ExpertPack. Migrates the 5-layer memory system (SESSION-STATE hot RAM, LanceDB warm store, Git-Notes cold store, MEMORY.md curated archive, and daily journals) into ExpertPack's portable format with multi-layer retrieval, context tiers, and EK measurement. Output is Obsidian-compatible — includes YAML frontmatter on all content files and can be opened as an Obsidian vault. Use when: upgrading from Elite Longterm Memory to ExpertPack, backing up agent knowledge, or migrating to a new platform. Triggers on: 'elite to expertpack', 'convert elite memory', 'export elite memory', 'migrate elite longterm', 'upgrade memory to expertpack', 'elite memory export'.

SKILL.md

2.8 KB, 522 tokens by cl100k_base, as published. Nobody here has run it

Elite Longterm Memory → ExpertPack

Converts an Elite Longterm Memory (5-layer system with 32K ClawHub downloads) into a proper structured ExpertPack.

Supported layers:

Hot RAM — SESSION-STATE.md (current task, context, decisions)
Warm Store — LanceDB vectors at ~/.openclaw/memory/lancedb/ (note: exported or skipped)
Cold Store — Git-Notes JSONL (decisions, learnings, preferences)
Curated Archive — MEMORY.md, memory/YYYY-MM-DD.md journals, memory/topics/*.md
Cloud — SuperMemory/Mem0 (skipped, noted in overview)

Usage

cd /root/.openclaw/workspace/ExpertPack/skills/elite-to-expertpack
python3 scripts/convert.py \
  --workspace /path/to/your/workspace \
  --output ~/expertpacks/my-agent-pack \
  [--name "My Agent's Knowledge"] \
  [--type auto|person|agent]

Flags let you override auto-detected paths for each layer.

What It Produces

A complete ExpertPack conforming to schema 2.3:

manifest.yaml (with context tiers, EK stub)
overview.md summarizing conversion (layer counts, warnings)
Structured directories: mind/, facts/, summaries/, operational/, relationships/, etc.
_index.md files, lead summaries, glossary.md (if terms found)
relations.yaml (if relationships detected)
Clean deduplication preferring curated > structured > raw sources

Secrets are automatically stripped (sk-, ghp_, tokens, passwords). Warnings emitted for any found.

Post-Conversion Steps

cd ~/expertpacks/my-agent-pack
Verify content files are 400–800 tokens each (Schema 2.5 — retrieval-ready by design)
Measure EK ratio: python3 /path/to/expertpack/tools/eval-ek.py .
Review overview.md and manifest.yaml
Commit to git and publish to ClawHub

Learn more: https://expertpack.ai • ClawHub expertpack skill

See also: Elite Longterm Memory skill on ClawHub.