observe

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt directs the agent to read the user's config (cat ~/.config/observe.yaml) and shows examples of passing auth tokens and passwords on the CLI/flags (e.g., --authtoken, direct password), which requires handling and potentially outputting secret values verbatim.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The README instructs installing the skill by curling the skill document from https://raw.githubusercontent.com/observeinc/observeinc-observe-claude-code-skill-example/main/SKILL.md which would fetch external content that directly defines/controls the agent's prompts/instructions and is required for the skill to function.