The Agent Skills Directory

[PROMPT_INJECTION]: The skill tracks progress by reading GitHub issue comments and pull request metadata, which creates an indirect prompt injection surface for data from external contributors. This risk is effectively mitigated by 'Hard-wait' and 'Hard gate' checkpoints that require manual user approval before the agent creates issues, pushes code, or closes tasks. 1. Ingestion points: gh issue view and gh pr list in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: GitHub CLI operations, file system modifications, and pnpm build command execution. 4. Sanitization: Absent.
[COMMAND_EXECUTION]: The skill uses standard development tools, specifically the GitHub CLI (gh) and the pnpm package manager, to perform tasks within the wso2/labs-agentic-engineer repository. These actions are standard for the development workflow and are performed within the vendor's own infrastructure.

console-feature