Tube Cog
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.
Installing or using the skill may require giving the agent access to a CellCog account/API credential.
The skill requires a CellCog API key, which is expected for a CellCog-powered content-generation service but is still a sensitive credential.
requires:\n bins: [python3]\n env: [CELLCOG_API_KEY]
Use a dedicated, least-privileged CellCog API key if available, and revoke or rotate it if you stop using the skill.
The reviewed artifact mainly provides instructions, so actual runtime behavior depends on the separate CellCog SDK or skill installed in the environment.
The skill relies on an external CellCog dependency/skill for the actual SDK behavior; that dependency is central to the purpose but not included as code in this artifact.
dependencies: [cellcog]
Install CellCog only from a trusted source and review its permissions and documentation before providing credentials.
Prompts, creative briefs, and any assets or instructions included in the task may be sent to CellCog for processing.
The documented workflow sends the user's task prompt to a CellCog chat/agent session and uses a notification session key; this is purpose-aligned but involves an external agent/provider boundary.
result = client.create_chat(\n prompt="[your task prompt]",\n notify_session_key="agent:main:main",\n task_label="my-task",\n chat_mode="agent",\n)
Avoid sending private or confidential material unless you trust CellCog’s handling of that data and its retention policy.
A started generation job may run asynchronously through CellCog rather than completing entirely in the current chat turn.
The OpenClaw example describes a non-blocking external task flow. This appears aligned with video generation, but users should understand tasks may continue after being started.
**OpenClaw (fire-and-forget):**
Start only clearly scoped tasks, label them meaningfully, and monitor CellCog/OpenClaw task results before using or publishing generated content.