The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using child_process.execSync in src/lib.ts to analyze git contribution history.
Evidence: The command git log --format="%an" --since="${since}" | sort | uniq -c | sort -rn is executed.
Context: The ${since} variable is sourced from an external configuration file (rules.json) loaded via KnowledgeProvider. The command is executed within the context of a user-provided directory (argv.dir), which is passed as the cwd option.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted external data (source code files) and possesses powerful capabilities.
Ingestion points: The skill reads source files (.js, .ts, .py) from the target directory in src/lib.ts (assessCodeQuality).
Boundary markers: No boundary markers or "ignore" instructions are used when reading and processing file content.
Capability inventory: The skill can execute shell commands (execSync), read arbitrary files in the target directory (fs.readFileSync), and write reports to disk (safeWriteFile).
Sanitization: No sanitization or validation of the ingested code content is performed before processing or including it in the logic flow.

tech-dd-analyst