The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands via bash to interact with git, gh (GitHub CLI), and glab (GitLab CLI). This is used for retrieving local project history and fetching repository data from GitHub or GitLab.
[PROMPT_INJECTION]: The skill has a vulnerability surface for indirect prompt injection because it ingests untrusted data from external sources and project metadata.
Ingestion points: Fetches issue titles, PR descriptions, and repository documentation from external repositories using CLI tools and the browse skill as defined in references/competitive-scan.md.
Boundary markers: The instructions do not specify the use of unique delimiters or XML tags when interpolating untrusted external content into the agent's context.
Capability inventory: The skill has the ability to execute shell commands via bash and perform file writes to ROADMAP.md and .gitignore.
Sanitization: The skill mitigates risks through an 'AI Self-Check' that mandates verification of claims against primary documentation and checks for hallucinated competitive data.
[EXTERNAL_DOWNLOADS]: The skill fetches repository information and documentation from well-known services (GitHub and GitLab) to perform competitive intelligence scans. These operations are conducted using official CLI tools or web fetching.

roadmap