self-discovery

Resolves and loads a canonical skill file from a verified repository path.

Updated Mar 22, 2026
One-click install
npx skills add https://github.com/diazMelgarejo/orama-system --skill self-discovery-diazmelgarejo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: self-discovery
Source: https://github.com/diazMelgarejo/orama-system/tree/main/.agents/skills/self-discovery
Command: npx skills add https://github.com/diazMelgarejo/orama-system --skill self-discovery-diazmelgarejo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? It prevents stale or duplicated skill definitions by acting as a thin wrapper that locates and loads the single canonical SKILL.md from the orama-system repository instead of maintaining local copies. ## Core Features & Use Cases - Marker-Verified Resolution: Searches environment variables, the current git repository root, and sibling checkouts for the canonical skill file, never guessing or falling back to unverified copies. - Read-Only Loading: Reads the canonical SKILL.md without fetching, pulling, installing, or modifying anything during skill activation. - Explicit Refresh Workflow: Supports a separate, explicitly authorized git fetch and fast-forward pull for maintenance, with dirty-worktree drift reporting. - Use Case: When multiple projects reference the same orama-system skill, this wrapper ensures every agent loads the authoritative version from the canonical checkout rather than a divergent local copy. ## Quick Start Ask the agent to load the self-discovery skill so it resolves the canonical orama-system repository and follows the authoritative skill instructions.

Frequently Asked Questions about self-discovery

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I resolve a canonical skill file across multiple git repositories?

Check the ORAMA_SYSTEM_ROOT or ORAMA_SYSTEM_PATH environment variables first, then the current git repository root, then a bounded search of parent and grandparent directories for a sibling checkout containing the marker file. Never hardcode workstation paths.

What is a thin wrapper skill in an agent skill system?

A thin wrapper skill contains no operational logic itself and instead points to a canonical SKILL.md in another repository. It resolves the canonical path at load time and follows those instructions, avoiding duplicated or divergent skill definitions.

Does loading this skill modify the canonical repository?

No. Loading is strictly read-only and marker-verified. Fetching, pulling, pruning, or installing only happens during an explicitly authorized refresh, which uses fast-forward-only pulls and refuses to overwrite dirty worktrees.

What happens if the canonical skill path cannot be found?

The wrapper reports the canonical skill as unavailable and asks for its location only if the task genuinely needs it. It never guesses, hardcodes a path, or falls back to a different repository's copy.

How do I handle UTF-8 encoding for skill files on Windows PowerShell?

Set the console input and output encodings to UTF-8, assign $OutputEncoding to a UTF8Encoding instance, and set the PYTHONUTF8 environment variable to 1 before reading or writing skill files.