Memory Trace
v1.0.4从小说、剧本、动漫素材中提取角色人格，生成 SoulPod 包。下游配合 Memory-Inhabit 加载对话。
⭐ 0· 110·1 current·1 all-time
byAsuka@evangeliona
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (persona extraction → SoulPod) matches included scripts (analyzer.py, forge.py) which parse text/PDF, infer personality, extract memories, and generate output/<character>/ packages. The forge script explicitly integrates with a sibling Memory-Inhabit persona directory, which is consistent with the described downstream workflow.
Instruction Scope
SKILL.md and README instruct running local Python scripts and installing pdfplumber. Runtime instructions and scripts only read provided input files, parse content, generate JSON/text outputs under output/, and optionally copy packages into Memory-Inhabit/personas. There are no instructions to read unrelated system files, environment secrets, or to send data to external endpoints.
Install Mechanism
No install spec is provided (instruction-only); code files are bundled with the skill. The scripts use only standard libraries and an optional pdfplumber dependency. There are no downloads from remote URLs or archive extraction steps in the install process.
Credentials
The skill declares no required environment variables, no credentials, and the code does not access secret-named env vars. All file I/O is local and scoped to source files, output/, and the target Memory-Inhabit/personas path; requested access is proportionate to the task.
Persistence & Privilege
always is false. The forge.py install action copies generated pods into a Memory-Inhabit/personas directory (sibling path) which is expected for integration, but it will attempt to write into that location if run. Autonomous invocation (model invocation enabled) is normal and not accompanied by other red flags.
Assessment
This skill appears coherent and implements exactly what it claims: local text/PDF parsing, persona modelling, and package generation. Before installing or running it: 1) Inspect forge.py's install routine (it will copy files into a Memory-Inhabit/personas path) and confirm you are comfortable granting write access to that directory. 2) Be aware it processes any text you give it — don't feed it sensitive secrets or proprietary material you don't want duplicated. 3) The package origin is 'unknown' and there's no homepage; if provenance matters, prefer code from trusted sources or review the full scripts yourself. 4) It depends on pdfplumber for PDFs (pip install), so install that in an isolated environment if needed. Otherwise the skill is internally consistent and does not request credentials or perform network exfiltration.Like a lobster shell, security has layers — review code before you run it.
latestvk97f5kgeqg86kn9q95yyfhv8nx84w853
License
MIT-0
Free to use, modify, and redistribute. No attribution required.