The Agent Skills Directory

[SAFE]: The skill is designed to manage security journals for Bitrix modules. It operates entirely on local files and uses standard Python libraries for JSON processing and data validation.\n- [DATA_EXFILTRATION]: The skill performs local file read and write operations on security-journal.json and reviewed-scan.json. No network operations or external data transmissions are present.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes external scan data, creating a potential surface for indirect prompt injection. Mandatory Evidence Chain: 1. Ingestion points: reviewed-scan.json and security-journal.json in SKILL.md and scripts/update_journal.py. 2. Boundary markers: The skill uses deterministic script logic and strict JSON schema validation rather than relying on natural language instructions for processing. 3. Capability inventory: Local file read/write and JSON data manipulation. 4. Sanitization: The script implements validate_scan and validate_journal functions that enforce strict requirements on input data structures.

journal-update