Linkedin Thread Engagement
ReviewAudited by ClawScan on May 10, 2026.
Overview
The skill matches its LinkedIn monitoring purpose, but it may route replies through an auto-posting backend without clearly defined approval or account permissions.
Before installing, confirm that the active reply backend is manual or requires explicit approval for every LinkedIn reply or DM. Review any HarvestAPI, Publora, or linkedin-reply-handler configuration separately, and do not grant posting permissions unless you want the agent to publish on your behalf.
Findings (4)
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 could publish LinkedIn replies through a configured backend when the user expected only draft text.
A workflow framed as drafting can invoke a backend that posts publicly, but the instructions do not define an explicit user confirmation gate before high-impact posting.
Draft responses for warm threads using `linkedin-reply-handler` (which adapts to the active backend per `lib.active_backend()` — Publora auto-posts, manual mode returns copy-paste, DIY invokes custom poster).
Use this only with a manual/copy-paste backend unless the user explicitly approves each post; require a confirmation step before any public reply or DM is sent.
The user may not know which LinkedIn or publishing account permissions the workflow will use, or how broad those permissions are.
Auto-posting or custom posting implies use of delegated LinkedIn/social publishing authority, but the supplied requirements declare no primary credential, env vars, or scoped account configuration.
Publora auto-posts, manual mode returns copy-paste, DIY invokes custom poster
Clearly document required accounts, tokens, scopes, and which backend is active before enabling the skill; avoid granting posting permissions unless necessary.
Actual posting behavior may depend on another installed skill or backend with separate risks.
Core behavior depends on another skill and active backend logic that are not part of this reviewed package, so their safety controls cannot be verified from these artifacts.
`linkedin-reply-handler` — drafts the actual follow-up message
Review the linked reply-handler skill and backend configuration before relying on this workflow, especially if any backend can post automatically.
LinkedIn activity and profile/thread information may be retrieved through or shared with an external provider.
The skill discloses a third-party/API data flow involving the user's LinkedIn profile URL, recent comments, and thread context; this is purpose-aligned but should be visible to the user.
Fetch user's recent comments via HarvestAPI `/linkedin/profile-comments`.
Use only if the user is comfortable with HarvestAPI handling this LinkedIn activity, and avoid providing private or sensitive profile data unnecessarily.