The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to perform local system operations using standard utilities to validate project state and progress. These include:
Using grep to check for completion markers (TODO/FIXME) in documentation files.
Using jq and diff via process substitution to compare feature titles in feature_list.json against the next steps recorded in progress.md.
Using git log with the --grep flag to filter and verify the development history associated with specific feature names.
[INDIRECT_PROMPT_INJECTION]: The skill processes project-level data that could be influenced by external inputs (such as pull requests or shared files), creating an attack surface when this data is interpolated into shell commands.
Ingestion points: The skill reads project state and feature descriptions from feature_list.json and progress.md.
Boundary markers: The instructions lack specific boundary markers or "ignore embedded instructions" warnings for the data ingested from these local state files.
Capability inventory: The skill possesses the capability to execute shell utilities (grep, jq, diff, git) and to orchestrate the execution of 9 other upstream skills.
Sanitization: There are no explicit requirements or steps provided for sanitizing or escaping the data extracted from the project files before it is passed to the shell environment.

harness-orchestrator