The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from project files (e.g., .planning/phases/**/*-SUMMARY.md and VERIFICATION.md) to identify exports, API routes, and status. This information is interpolated into prompts for the kata-integration-checker subagent and used as arguments in shell commands (grep). Maliciously crafted content in these project files could theoretically influence subagent behavior or lead to command injection.
Ingestion points: SKILL.md reads verification and summary files across multiple phase directories. integration-checker-instructions.md instructs the subagent to extract key exports and routes from these files.
Boundary markers: The skill does not implement specific boundary markers or instructions to ignore embedded commands when interpolating extracted data into subagent tasks or shell scripts.
Capability inventory: The skill can execute shell commands (find, grep, cat, node) and spawn new subagents using the Task capability.
Sanitization: No explicit validation or escaping is performed on the extracted strings before they are used in prompts or shell commands.
[COMMAND_EXECUTION]: The skill executes local shell commands to discover project files, read phase reports, and analyze source code structure. It also utilizes a bundled Node.js script (scripts/kata-lib.cjs) to manage configuration settings within the project's .planning directory.
[DYNAMIC_EXECUTION]: The skill dynamically spawns a kata-integration-checker subagent to perform specialized integration and end-to-end flow verification, passing it instructions read from a local reference file and context gathered from the project environment.

kata-audit-milestone