Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Supercharged Daily Briefing
v1.0.3Stop spending your mornings hunting for news, trends, and updates across a dozen tabs. The Supercharged Daily Briefing turns OpenClaw into a production-grade...
⭐ 0· 80·0 current·0 all-time
by@nollio
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
The skill's name and description (discover sources, fetch feeds, synthesize briefings) match the instructions and files: SKILL.md, local JSON registries, and a scheduler script. One minor mismatch: the package metadata lists no required binaries, but scripts call python3 and standard POSIX tools (mkdir, chmod, grep, cp); that should be declared but is proportionate to the purpose.
Instruction Scope
Runtime instructions operate on local workspace files (config/, data/, scripts/) and on user-approved web URLs via web_search/web_fetch. The SKILL.md explicitly includes a prompt-injection defense and URL safety rules (only http/https, block localhost/private IPs). The setup prompt asks the agent to copy files into the workspace and set permissions — expected for installation, and the skill does not instruct reading unrelated user files or sending data to unknown endpoints.
Install Mechanism
No remote install/downloads or extracted archives are present; it's an instruction-only skill that copies provided files into the workspace. This is low risk. Note: there is no install spec but SETUP-PROMPT.md will copy scripts into place — everything stays local.
Credentials
The skill declares no environment variables, credentials, or external telemetry. Dashboard/add-on docs mention optional Supabase/Postgres if you choose to deploy a dashboard, but that is optional and not required by the core skill.
Persistence & Privilege
always:false and no requests to modify other skills or global agent settings. The setup copies files and changes permissions for its own workspace — normal for a local skill. Autonomous invocation remains allowed (platform default) but the skill does not request elevated or persistent cross-skill privileges.
Scan Findings in Context
[prompt-injection:ignore-previous-instructions] expected: The SKILL.md explicitly cites 'ignore previous instructions' and other command-like phrases as examples of prompt-injection to be ignored; the pre-scan matcher flagged the phrase but it appears intentionally included as a defense rather than an attack payload.
Assessment
This package appears coherent with its stated purpose, but take these simple precautions before installing:
- Inspect and (if you can) run the included setup validation: ./scripts/briefing-scheduler.sh --check. Note: the scheduler script has a broken/strange find_skill_root implementation — review/fix it before wide use.
- Ensure your environment has the expected utilities (bash shell, python3, grep, cp, chmod). The metadata didn't list python3 but the script invokes it.
- SKILL.md asks the agent to fetch external HTTP(S) sources. Confirm your agent's web_fetch/web_search implementations behave as you expect and that you only approve sources you trust.
- The skill stores data locally (data/ and config/). If you plan to enable the optional dashboard or sync to Supabase/Postgres, review those components separately — they may require externally hosted services.
- The pre-scan flagged a prompt-injection string; this skill includes explicit defenses against injection, but remain cautious: avoid installing in environments with sensitive host-local services exposed (the skill explicitly blocks localhost/private IPs when fetching, but it's good to verify your agent enforces that).
If you want higher assurance, run the setup steps in an isolated workspace, confirm file permissions, and manually fix the small script bug before enabling scheduled runs.SKILL.md:18
Prompt-injection style instruction pattern detected.
About static analysis
These patterns were detected by automated regex scanning. They may be normal for skills that integrate with external APIs. Check the VirusTotal and OpenClaw results above for context-aware analysis.Like a lobster shell, security has layers — review code before you run it.
latestvk97c6k2nk2hdcbjr07g7vd770x83ywew
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
