Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Fox Veille
v1.0.0RSS feed aggregator, deduplication engine, LLM scoring, and output dispatcher for OpenClaw agents. Use when: fetching recent articles from configured sources...
⭐ 0· 40·1 current·1 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (RSS aggregator, dedup, LLM scoring, dispatch) matches the provided code and runtime instructions. Files show feed fetching, TTL-based seen-store, topic dedup, optional LLM scoring (reads an API key file), and dispatchers for Telegram/email/Nextcloud/file—all appropriate for the stated purpose. Minor metadata mismatch: _meta.json ownerId/version differs from registry metadata, which is an administrative inconsistency but not a functional mismatch.
Instruction Scope
SKILL.md and CLI instruct only feed fetching, scoring, deduplication, and dispatch. The only cross-config read documented is ~/.openclaw/openclaw.json for a Telegram bot token when telegram output is enabled; scorer reads an API key file only when LLM scoring is enabled. No instructions asking to read unrelated secrets or broad system state. Scheduled/autonomous dispatch is documented (cron), which is expected for a digest skill.
Install Mechanism
No install spec (instruction-only); code is included in the skill bundle. There are no network downloads or arbitrary extract/install steps in the manifest. Running occurs locally via provided Python scripts, consistent with the project's description.
Credentials
The skill declares no required environment variables. It reads local config/data under ~/.openclaw and an optional LLM API key file (default ~/.openclaw/secrets/openai_api_key) only when scoring is enabled. The only cross-config read is the Telegram bot token from ~/.openclaw/openclaw.json when telegram output is enabled and no bot_token is provided—this behaviour is documented. Overall credential access is proportional to functionality.
Persistence & Privilege
always:false (no forced inclusion). The skill can create a scheduled job (cron) via setup.py and perform autonomous dispatch when scheduled; this is documented. Delegation to other skills uses subprocess.run() with path validation to ~/.openclaw/workspace/skills/, which limits but does not eliminate risk if other installed skills are malicious. No evidence the skill modifies other skills' configs or elevates privileges.
Scan Findings in Context
[pre-scan-none] expected: The provided pre-scan reported no injection signals. Manual review of the code snippets shows network calls, filesystem reads/writes, and subprocess delegation that are expected for this skill.
Assessment
This skill appears internally consistent with its stated purpose. Before installing: 1) Inspect setup.py (it can create a cron job) and confirm you want scheduled autonomous dispatch. 2) If you enable Telegram output and want to avoid any cross-config reads, set bot_token explicitly in the skill output config (SKILL.md shows how). 3) If you enable LLM scoring, place the API key in a dedicated file (e.g. ~/.openclaw/secrets/openai_api_key) and check its filesystem permissions. 4) Note the minor administrative mismatch between registry metadata and _meta.json (owner/version); verify the skill source/repo you install from. 5) Review any other installed skills before allowing this skill to delegate to them (dispatch validates script paths, but it will run scripts under ~/.openclaw/workspace/skills/). If you want extra caution, run python3 scripts/init.py to validate network access and python3 scripts/veille.py fetch --hours 24 in a non-production environment to observe behaviour before enabling outputs.scripts/dispatch.py:89
Dynamic code execution detected.
Patterns worth reviewing
These patterns may indicate risky behavior. Check the VirusTotal and OpenClaw results above for context-aware analysis before installing.Like a lobster shell, security has layers — review code before you run it.
latestvk976hh8zqjb17xssrqsttfmggs83w44c
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
📡 Clawdis