Skills

This repository owns Agentic Tend's reusable skills layer, dogfooded in Codex before publication.

Each top-level directory is one independently discoverable skill with a SKILL.md entrypoint. The collection follows the open Agent Skills format: compatible agents discover concise metadata first and load full instructions and bundled resources only when a task calls for them. Agentic Tend's tooling model keeps these on-demand workflows separate from ambient repository rules and event-triggered hooks.

Below separates meta-level procedures that shape agent work from adapters for a specific capability.

Validate

Validate every published skill against the Agent Skills specification:

gh skill publish --dry-run

Agent workflow skills

These meta-level skills govern how an agent reasons about and structures its work across repositories.

• Clarifying Contracts resolves user-owned ambiguity about purpose, observable behavior, constraints, scope, authority, and acceptance evidence before execution. It adapts Superpowers' brainstorming workflow and Matt Pocock's grilling¹², focusing on the intrinsically complex entries of HITL (human-in-the-loop).

Functional skills

These skills guide an agent in using a particular tool or capability safely and consistently.

• Obsidian CLI adapts and narrows Kepano's upstream Obsidian CLI skill, distinguishing non-CLI versus CLI processes and non-DEV versus DEV tasks.

See also

• The organization roadmap tracks evidence and extension work spanning multiple tooling layers.

Footnotes

1. obra, Superpowers: composable, enforceable workflows for agentic software development. ↩

2. Matt Pocock, grilling: one-question-at-a-time clarification of plans and decisions. ↩