The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes arbitrary task descriptions provided by the user, which are then passed into the context of secondary planning and implementation skills.
Ingestion points: The <task description> and plan file paths ingested in SKILL.md and README.md.
Boundary markers: Absent; the skill does not explicitly wrap or delimit user-provided task descriptions before forwarding them to sub-skills.
Capability inventory: The orchestration involves triggering gitnexus-work, which possesses the capability to perform filesystem writes and create git commits.
Sanitization: No sanitization or validation of the input task string is performed by this orchestrator.
[EXTERNAL_DOWNLOADS]: The skill's configuration downloads and executes a specific tool from an external package registry.
Evidence: The mcp.json file specifies the command npx -y gitnexus@1.6.12 mcp to fetch and run the gitnexus package.

gitnexus-lfg