The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill requires the agent to process feedback from the Fusion 360 environment via a local file (results.json). Ingestion points: results.json (SKILL.md). Boundary markers: Absent. Capability inventory: The agent has file-write access to commands.json (SKILL.md). Sanitization: Absent. This surface could allow an attacker to embed instructions in CAD metadata or design objects that override the agent's logic.
[EXTERNAL_DOWNLOADS]: The skill's documentation specifies a mandatory dependency on the ClaudeBridge add-in, directing users to a third-party repository (altjx/ClaudeBridge) on GitHub.
[DATA_EXFILTRATION]: The inclusion of an export_session command allows for the bulk collection and export of design data to a local directory, providing a mechanism for sensitive intellectual property to be aggregated.

fusion360