The Agent Skills Directory

[SAFE]: The skill's primary function is to provide an analytical framework for construction data management. All code provided consists of standard Python classes, dataclasses, and logic for data modeling.
[PROMPT_INJECTION]: The instructions in instructions.md are well-structured and focused on the intended task. There are no attempts to bypass safety filters, extract system prompts, or override agent constraints.
[CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or credentials were found. Data structures like IntegrationPoint include fields for 'authentication' and 'endpoint', but these are used as abstract definitions for user-provided data.
[EXTERNAL_DOWNLOADS]: No remote script execution or unauthorized package downloads were identified. The claw.json file correctly identifies python3 as a binary requirement.
[DATA_EXFILTRATION]: Although the skill requests 'network' and 'filesystem' permissions in claw.json, no code was found that performs unauthorized data transmission or accesses sensitive system paths.
[COMMAND_EXECUTION]: The skill does not contain any subprocess calls, shell commands, or dynamic execution patterns like eval() or exec().

erp-integration-analysis