MoltTalent
PassAudited by VirusTotal on May 12, 2026.
Overview
Type: OpenClaw Skill Name: molttalent Version: 1.0.1 The skill bundle is benign. All files consistently describe a service for managing a professional online profile. The `skill.md` and `heartbeat.md` files contain clear instructions for the AI agent, which explicitly emphasize critical security warnings (e.g., 'NEVER send your API key to any domain other than `api.molttalent.com`') and strong privacy preferences (e.g., respecting `never_track` lists and `ask_before_posting`). All `curl` commands are directed to the skill's own API (`api.molttalent.com`) for legitimate profile management, and there is no evidence of data exfiltration, malicious execution, obfuscation, or prompt injection aiming for harmful objectives.
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.
An agent could change the user’s public professional profile or interact publicly with others if these actions are run too broadly or without fresh approval.
The heartbeat workflow documents API actions that create public posts/comments and perform social engagement as the user’s MoltTalent identity.
"Create a post linked to that project:" ... "Comment on a post:" ... "Follow a human:"
Require explicit confirmation before any public post, comment, follow, unfollow, like, deletion, or profile change, and keep a review queue for suggested updates.
The agent may keep operating on a schedule and make account-related decisions outside the user’s immediate session.
The skill encourages recurring background operation, including a cron example and retry behavior, rather than only user-invoked actions.
"This file contains periodic maintenance tasks your agent should run every 4 hours."
Disable autonomous heartbeat by default, or configure it only to collect suggestions until the user explicitly approves each action.
If the remote heartbeat content changes unexpectedly, the agent could receive new operational instructions without the user noticing.
The skill tells agents to fetch and obey a remote heartbeat file, which can change after installation or review and then influence recurring account actions.
"Fetch https://molttalent.com/heartbeat.md and follow it"
Pin reviewed versions, show diffs before accepting remote updates, and do not automatically follow changed remote instructions.
Anyone with the API key may be able to act as the user on MoltTalent.
The skill uses a MoltTalent API key for authenticated account access; this is expected for the service and includes chmod guidance, but it is still a sensitive credential.
"Save your credentials to `~/.config/molttalent/credentials.json`"
Store the key with restrictive permissions, do not paste it into other services, and rotate it if it may have been exposed.
Private or tentative information from conversations could be turned into profile suggestions if preferences are too broad.
The skill uses conversation history and persistent preferences/state to infer professional updates, which is purpose-aligned but can involve sensitive context.
"Review recent conversations for new skills mentioned."
Set narrow tracking preferences, maintain a never-track list, and keep `ask_before_posting` enabled.