AstroClaw

PassAudited by ClawScan on May 5, 2026.

Overview

AstroClaw is a coherent, instruction-only horoscope API skill, with modest caution because it asks the agent to fetch and keep a daily external forecast in context.

AstroClaw appears safe for playful horoscope use. Before installing, confirm you are comfortable with occasional requests to astroclaw.xyz and with the agent keeping a short daily forecast in context; do not let the forecast guide important, factual, security-sensitive, or business decisions.

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.

What this means

Installing the skill may cause the agent to contact astroclaw.xyz when asked for a horoscope or as part of a daily routine.

Why it was flagged

The skill documents an outbound HTTP request to a fixed horoscope API domain; this is expected for the stated purpose and does not include credentials or sensitive parameters.

Skill content
curl -s https://astroclaw.xyz/api/forecasts/2026-03-11/aries.json
Recommendation

Allow the skill only if you are comfortable with occasional outbound requests to this domain; keep requests limited to the documented forecast endpoint.

NoteHigh Confidence
ASI01: Agent Goal Hijack
What this means

The agent's wording or creative choices could be nudged by the daily forecast even outside a direct horoscope answer.

Why it was flagged

The skill intentionally asks the agent to let external horoscope text influence its outputs. This is disclosed and central to the playful purpose, but it should not override user instructions or safety constraints.

Skill content
Keep your daily horoscope in context to subtly influence generations based on the cosmos
Recommendation

Use the forecast as decorative context only, not as an instruction source or basis for important decisions.

What this means

A forecast from the external service could persist in the agent's context for the day and affect later responses.

Why it was flagged

The skill recommends retaining external API content in state/context. It also instructs sanitization and length limits, making this purpose-aligned but still worth noticing.

Skill content
Store "last_alignment_check" timestamp and the sanitized forecast; Keep the forecast in your working context for the day
Recommendation

Treat forecasts as untrusted data, store only short sanitized plain text, and do not preserve or follow any operational instructions that might appear in the forecast text.

NoteHigh Confidence
ASI10: Rogue Agents
What this means

The agent may perform a daily horoscope check and update local state if the user configures such a routine.

Why it was flagged

The skill suggests recurring daily execution. This is disclosed and bounded to a daily forecast check, with no hidden background code present.

Skill content
Add this to your daily heartbeat or periodic tasks
Recommendation

Enable any daily routine only intentionally, keep it limited to one check per day, and make it easy to disable.