LinkedIn Follow-up
AdvisoryAudited by Static analysis on Apr 30, 2026.
Overview
No suspicious patterns detected.
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.
The agent may act through your Google and LinkedIn accounts, including reading private messages, updating spreadsheets, and sending LinkedIn messages as you.
The skill uses Google authorization tokens and an authenticated LinkedIn browser session, while the supplied registry metadata declares no primary credential or required config path.
`gog auth` — Run `gog auth list`... `Browser` — Open the openclaw browser profile and confirm LinkedIn is logged in.
Use a dedicated browser profile and least-privileged Google Sheet, confirm the Sheet ID is yours, and require explicit approval before every send or account mutation.
A wrong draft, wrong recipient, or accidental batch operation could send real messages under your LinkedIn identity.
The browser workflow can directly click LinkedIn’s real Send button from the logged-in profile. This matches the skill purpose, but it is an irreversible external account action.
### 4d. Send ... if (btn) { btn.click(); return 'sent'; }Keep the approval step mandatory, verify screenshots and recipients before sending, and avoid unattended or bulk sends.
Using stealth-like LinkedIn automation may put the user’s account, reputation, or compliance posture at risk.
The workflow explicitly includes anti-detection behavior for LinkedIn automation without explaining platform, account, or compliance risks to the user.
**Always go to feed first** (anti-detection): `https://www.linkedin.com/feed/`
Remove anti-detection guidance, disclose platform-risk tradeoffs clearly, and prefer official or policy-compliant integration methods.
Private conversation history may be retained and reused beyond the current task, and anyone with sheet access could see it.
The skill stores LinkedIn conversation content as persistent CRM context for later use, which can expose private messages in the sheet and allow untrusted message text to influence future drafts.
Append new entries to the existing log in col O... Keep the format consistent so the follow-up skill can parse it in future sessions.
Restrict sheet access, confirm the destination Sheet ID, redact sensitive content, set a retention policy, and treat stored conversation text as untrusted input.