Back to skill
Skillv1.2.1
ClawScan security
Publora · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
SuspiciousMar 20, 2026, 1:00 PM
- Verdict
- suspicious
- Confidence
- medium
- Model
- gpt-5-mini
- Summary
- The skill's instructions match a social-posting API, but metadata omissions and a credential declaration mismatch leave unanswered questions you should resolve before using it.
- Guidance
- This SKILL.md appears to be a straightforward client for the Publora REST API, but before installing or using it you should: 1) verify the service and endpoints by visiting publora.com (the skill metadata has no homepage or source URL); 2) confirm how/where you will supply your x-publora-key — avoid pasting long-lived keys into prompts if you don't trust the skill or agent UI; 3) prefer creating a limited/test API key (or sandbox account) with minimal permissions for initial testing; 4) check Publora's privacy and webhook behavior (webhooks or media uploads may expose content/URLs externally); and 5) if the agent asks to store the key, verify storage policy and delete any test keys afterward. The main issues are the missing source/homepage and the fact the skill does not declare the required API credential — both are legitimate reasons to pause and validate provenance before proceeding.
Review Dimensions
- Purpose & Capability
- noteSKILL.md documents Publora endpoints for posting, scheduling, media uploads, webhooks, and analytics which align with the declared purpose (multi-platform social publishing). However the package metadata provides no source or homepage, and the skill does not declare any required API key even though every request in the instructions requires an x-publora-key. This is an inconsistency worth noting.
- Instruction Scope
- okInstructions are focused on calling publora.com API endpoints (platform-connections, create-post, list-posts, get-upload-url, etc.). They do not instruct the agent to read local files, environment variables, or unrelated system state, nor do they direct data to unexpected external endpoints beyond publora.com and pre-signed upload URLs.
- Install Mechanism
- okNo install spec and no code files (instruction-only). This minimizes filesystem risk because nothing is downloaded or written by the skill itself.
- Credentials
- concernThe runtime examples require an API key sent in the x-publora-key header (keys beginning with sk_). Yet the skill metadata declares no required environment variables or primary credential. The skill will therefore rely on the agent or user to supply secrets ad hoc; that mismatch is a proportionality/information gap and a potential usability/security risk (users may be prompted to paste secrets without a clear storage policy).
- Persistence & Privilege
- okalways is false and there are no install actions. The skill does not request persistent privileges or modify other skills or system config per the provided files.