The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/validate_module.py uses the eval() function to parse the contents of the __manifest__.py file. This is a significant security risk because eval() can execute arbitrary Python code. If the script is used to validate a module containing a maliciously crafted manifest file, it could result in unauthorized code execution on the user's system. It is highly recommended to use ast.literal_eval() instead, which is designed for safely evaluating strings containing Python literals.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external Odoo module data (manifests, views, models). There are no specific boundary markers or instructions to the agent to disregard potential instructions embedded within these files. Given that the skill also possesses capabilities like file writing and dynamic execution via the validation script, this creates a surface for indirect prompt injection attacks. Evidence chain:
Ingestion points: The agent is instructed to read Odoo module files; scripts/validate_module.py specifically reads __manifest__.py.
Boundary markers: Absent.
Capability inventory: The skill provides tools that perform file I/O (init_oca_module.py) and dynamic execution (validate_module.py).
Sanitization: Absent.
[PROMPT_INJECTION]: A heuristic detector flagged the mention of "developer mode" in the assets/module_template/readme/CONFIGURE.md file. However, analysis confirms this is a standard reference to a built-in Odoo administrative feature intended for end-user documentation and does not constitute a malicious attempt to bypass agent safety protocols.

odoo-oca-developer