每日新闻搜索与智能摘要

PassAudited by VirusTotal on May 10, 2026.

Overview

Type: OpenClaw Skill Name: news-digest-v1 Version: 1.0.2 The skill bundle exhibits high-risk behavior by programmatically accessing the OpenClaw configuration file (`~\.openclaw\openclaw.json`) in `scripts/news_digest_v2/stage2_5_llm_summary.py` to extract sensitive API keys and base URLs. While this is ostensibly for the stated purpose of news summarization, direct access to secret stores is a significant security risk. Furthermore, the script contains unusual coding patterns, such as using `__import__('datetime')` for a module already imported and referencing a likely non-existent model version (`qwen3.6-plus`). Although the bundle includes extensive and legitimate-looking news filtering logic in `rules_config.py`, the automated extraction of credentials warrants a suspicious classification.

Findings (0)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

A manipulated network response could lead to inaccurate or malicious-looking content being included in the digest.

Why it was flagged

After an SSL error, the scraper retries with TLS certificate verification disabled, which can let a network attacker alter fetched news pages.

Skill content
response = requests.get(url, headers=HEADERS, timeout=timeout, verify=False)
Recommendation

Prefer sources with valid HTTPS, remove or gate the verify=False fallback, and treat scraped content as untrusted.

What this means

Future package changes or a compromised dependency source could affect the scripts the user runs.

Why it was flagged

The documented setup installs unpinned Python packages; this is common and purpose-aligned, but package versions and provenance are not locked.

Skill content
pip install requests beautifulsoup4
Recommendation

Install in a virtual environment and consider pinning known-good dependency versions.

What this means

If enabled, the LLM provider key should be treated as sensitive account access.

Why it was flagged

The skill can use an optional LLM API key for summarization; this credential use is disclosed and aligned with the optional LLM feature.

Skill content
`NEWS_DIGEST_LLM_API_KEY` | (empty) | LLM API key for Stage 2.5 summarization
Recommendation

Set the API key only when needed, store it securely, and use a key scoped to this purpose if possible.

What this means

Public news content is normally low sensitivity, but customized private or internal sources could be shared with an external LLM provider.

Why it was flagged

When the optional LLM stage is enabled, scraped article content may be sent to the configured LLM API/provider.

Skill content
Stage 2.5: LLM → Batch LLM summarization (optional, requires API key)
Recommendation

Use a trusted LLM endpoint and avoid enabling LLM summarization for private sources unless that data sharing is acceptable.

NoteHigh Confidence
ASI10: Rogue Agents
What this means

If the user configures the cron job, the scraper can run unattended and produce or forward digests daily.

Why it was flagged

The skill documents an optional daily scheduled run; this is disclosed automation rather than hidden persistence.

Skill content
schedule: "0 20 * * *"  # Daily 20:00 ... run: python scripts/news_digest_v2/run_all_stages.py
Recommendation

Enable scheduling only intentionally, and review the destination and contents before automatic sharing.