The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple shell commands using git, gh CLI, yq, and jq to analyze and update project state. It also utilizes dynamic context injection to run a local configuration script (discover-config.sh) at load time, which is consistent with its administrative purpose.- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests and processes untrusted data from external sources.
Ingestion points: Recent git commit messages, GitHub/Trello issue titles and bodies, and project source code files.
Boundary markers: The skill does not implement explicit delimiters or boundary markers when interpolating this external data into its reasoning process.
Capability inventory: The skill has the ability to execute arbitrary bash commands, write to the file system, and perform network operations through integrated CLI tools.
Sanitization: No sanitization or validation of the ingested external content is performed before it is used to suggest actions like closing issues or generating ADRs. However, the risk is mitigated by the requirement for user confirmation before executing proposed changes.

reconcile