Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Meme Token Analyzer
v1.0.0Meme Token Analyzer workflow with web search, image generation, data cleaning, and multimodal analysis to output wealth gene detection reports. Use this skil...
⭐ 1· 57·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
Capability signals
These labels describe what authority the skill may exercise. They are separate from suspicious or malicious moderation verdicts.
OpenClaw
Suspicious
medium confidencePurpose & Capability
The SKILL.md and many code files implement a web-search → clean → image-gen → multimodal-analysis workflow (coherent). However registry metadata at the top claimed 'Required env vars: none' while SKILL.md declares a required COZE_WORKSPACE_PATH env var. The package contains S3 storage and database modules (src/storage/s3/s3_storage.py, src/storage/database/*) even though SKILL.md repeatedly states 'stateless operation' and 'no API keys in code'. Presence of storage and DB modules is disproportionate to a simple analysis skill unless optional persistence is intended — that intent is not clearly explained.
Instruction Scope
The runtime instructions require calling external web-search, image-generation, and LLM APIs (expected). But SKILL.md instructs users to 'MUST read the SDK guide' and contains install/run examples that will cause the code to execute network calls and possibly read configuration from a workspace path. The codebase includes filesystem utilities and S3 storage which suggests the skill may read/write files or persist results; SKILL.md's claim of 'stateless operation' conflicts with these files. The SKILL.md and code also reference a private SDK (coze_coding_dev_sdk) and a multimodal model (doubao-seed-1-6-vision-250815) which will require runtime credentials/context not declared in top-level registry metadata.
Install Mechanism
No platform-level install spec in the registry summary, but SKILL.md contains metadata.install instructing 'pip install -r requirements.txt'. Installing from requirements.txt is a standard mechanism but pulls packages (some private-looking: coze_coding_dev_sdk, langgraph, coze_coding_utils). These packages are traceable but require reviewing requirements.txt for unexpected third-party packages. No direct single-URL downloads or archives were detected in the manifest. Recommendation: inspect requirements.txt before pip install and prefer a sandboxed environment.
Credentials
Top-level metadata shown to the scanner earlier said 'Required env vars: none', but SKILL.md metadata declares COZE_WORKSPACE_PATH as required. The code uses runtime context (coze_coding_dev_sdk clients) that typically rely on external credentials or platform-provided auth; those credential requirements are not declared in the registry summary. Additionally, S3 storage code exists (suggesting potential AWS credential usage) but no AWS keys are declared or explained. The mismatch between declared and actual env/credential needs is a red flag — you should verify which env vars and secrets the runtime SDKs expect before providing credentials.
Persistence & Privilege
always:false (no forced inclusion) and disable-model-invocation:false (normal). The repository includes modules for persistent storage (database and s3) and scripts for local runs and packaging; these imply optional persistence that could store or upload data. SKILL.md claims stateless operation and 'no local API key storage', but code files for storage are present. This combination increases blast radius if the skill is allowed to run autonomously without auditing configuration for where it writes data.
What to consider before installing
What to check before installing:
1) Env/credentials: SKILL.md requires COZE_WORKSPACE_PATH but the registry metadata omitted env vars — confirm which environment variables and API credentials the SDKs (coze_coding_dev_sdk, image/LLM/search clients) will need. Do NOT provide cloud or API credentials until you verify where they are used.
2) Inspect requirements.txt and dependencies: review all third-party packages (especially any private SDKs) for expected behavior and whether they contact external endpoints or require secrets.
3) Review storage code: the repo contains S3 and database modules. If you don't want persistence or uploads, locate and audit src/storage/s3/s3_storage.py and any DB code to ensure they are disabled or safe. Ensure COZE_WORKSPACE_PATH does not point to sensitive locations.
4) Run in isolation: if you decide to test, run pip install in an isolated virtualenv/container with no production credentials, and run the skill with network access restricted if possible to observe outgoing endpoints.
5) Verify documentation claims: the included 'Final Review' and other reports assert 'stateless' and 'no API keys', but code presence contradicts that — treat these claims skeptically and confirm actual runtime behavior.
6) If you must provide credentials: use least-privilege keys, temporary tokens, and monitor outgoing requests. Prefer to supply only the minimal SDK credentials required (and never broad AWS keys unless necessary).
If you want, I can: (a) list and summarize contents of requirements.txt, (b) inspect s3_storage.py and any places that call os.getenv to enumerate undeclared env vars, or (c) scan the code for network endpoints and secret usages. Which would you like me to do next?Like a lobster shell, security has layers — review code before you run it.
latestvk97eq9k89agv9f2hftc6jvvg2984q67h
License
MIT-0
Free to use, modify, and redistribute. No attribution required.