The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected. The skill is purpose-built for documenting software verification results and follows standard pipeline practices.
[INDIRECT_PROMPT_INJECTION]: The skill ingests external implementation data and test results, which could potentially contain hidden instructions. However, the skill lacks high-risk capabilities like network access or privilege escalation that could be exploited.
Ingestion points: Implementation source code and test suite output processed during verification phases (Mode 2 and 3).
Boundary markers: No specific delimiters or warnings for embedded instructions are used.
Capability inventory: Artifact generation via markdown templates and report formatting via a local shell script.
Sanitization: Basic string formatting (replacing newlines) is performed in the reporting script to preserve markdown layout, though no full semantic sanitization is implemented.
[COMMAND_EXECUTION]: The skill includes a bash script (scripts/verification-report.sh) that executes a local Python one-liner for JSON parsing. All operations are local and limited to the defined reporting functionality.

sdd-verification