The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external data from GitHub that could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: Fetches GitHub issue content using gh issue view in SKILL.md to establish feature context and constraints.
Boundary markers: The instructions do not specify the use of delimiters or specific warnings to ignore instructions embedded within the fetched issue body during the council evaluation phase.
Capability inventory: The skill has broad capabilities to write back to the repository and project board, including creating git commits, pushing to remote branches, creating pull requests, adding issue comments, and updating project field values.
Sanitization: There is no explicit sanitization or filtering of the fetched issue content before it is processed by the AI councils (Product Council and Feature Council).
[COMMAND_EXECUTION]: The skill uses various command-line tools to interact with the project repository and infrastructure.
It executes git for branch creation and version control operations.
It utilizes the GitHub CLI (gh) extensively for issue management, project board updates, and pull request creation.
It runs pnpm exec prettier to ensure formatting consistency for the generated decision records and documentation.
[DYNAMIC_EXECUTION]: The skill uses inline script execution for data processing.
It employs python3 -c to parse and filter JSON output from the GitHub CLI when checking for existing project items and extracting identifiers.

plan-feature