Postiz Extended

PassAudited by VirusTotal on May 12, 2026.

Overview

Type: OpenClaw Skill Name: Developer: Version: Description: OpenClaw Agent Skill Suspicious High-Entropy/Eval files: 2 The skill bundle is classified as suspicious primarily due to the presence of hardcoded credentials (email: `sascha@mykuhlmann.com`, password: `Postiz2026!`) in both the `SKILL.md` instructions and the Python scripts (`scripts/check_duplicates.py`, `scripts/post.py`). While these credentials appear to be for a specific, likely demo, instance (`https://postiz.home.mykuhlmann.com`), hardcoding credentials in a skill bundle is a significant security anti-pattern that exposes sensitive information. The skill otherwise aligns with its stated purpose of managing social media posts, with all network and file system access (e.g., `/tmp/postiz-cookies.txt` for session management) being directly related to interacting with the Postiz API. There is no evidence of prompt injection against the agent, data exfiltration to unrelated endpoints, or other malicious behaviors.

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.

What this means

Using the skill could authenticate to and control someone else’s Postiz-connected social accounts, including publishing or managing posts.

Why it was flagged

The helper script embeds a real-looking Postiz host, email, and password instead of requiring a user-provided credential, granting account-level authority to the configured Postiz/social integrations.

Skill content
POSTIZ_URL = "https://postiz.home.mykuhlmann.com" ... CREDENTIALS = {"email": "sascha@mykuhlmann.com", "password": "[hardcoded]", "provider": "LOCAL"}
Recommendation

Do not use the embedded credentials. Rotate/remove the exposed password, require user-supplied secrets through environment variables or OAuth, and clearly declare the credential requirement in metadata.

What this means

A mistaken or unauthorized invocation could schedule, immediately publish, alter, or remove social posts.

Why it was flagged

Publishing, updating, and deleting social posts are expected for this skill, but they are high-impact actions because they affect public or business social media accounts.

Skill content
- List, query, update, and delete scheduled posts ... Post Types ... `now` — Publish immediately
Recommendation

Require explicit user confirmation before publishing, deleting, or updating posts, especially when using `now` or bulk/multi-platform actions.

What this means

A user may believe they are configuring their own Postiz account while the skill is actually prepared to operate against predefined accounts.

Why it was flagged

The description frames the skill as a general self-hosted/cloud Postiz integration, but the instructions point to specific named accounts and handles.

Skill content
Schedule and manage social media posts via Postiz API (self-hosted or cloud) ... Handle ... @CoolmannSa ... kuhlmannsascha ... coolmanns.bsky.social
Recommendation

Make the target Postiz instance and social accounts configurable, remove account-specific defaults, and clearly disclose any fixed account binding.

What this means

Running the helper may install or use a dependency that was not declared in the skill metadata.

Why it was flagged

The helper script relies on an inline, unpinned Python dependency when run with uv; this is common for small scripts but is not reflected in the registry requirements.

Skill content
# dependencies = ["requests"]
Recommendation

Pin dependencies and declare runtime requirements so users can review what will be installed or used.