_shared

Stores shared SDD reference documents for persistence conventions and phase protocols.

Updated Jul 10, 2026
One-click install
npx skills add https://github.com/AD-Paladins/beaconator-web --skill shared-ad-paladins
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: _shared
Source: https://github.com/AD-Paladins/beaconator-web/tree/main/.config/opencode/skills/_shared
Command: npx skills add https://github.com/AD-Paladins/beaconator-web --skill shared-ad-paladins

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? It centralizes the common reference documentation that all SDD (Spec-Driven Development) phase skills depend on, preventing duplication of persistence contracts, artifact naming conventions, and status schemas across every installed skill. ## Core Features & Use Cases - Engram Artifact Convention: Defines deterministic naming, two-step recovery (mem_search then mem_get_observation), and upsert rules for SDD artifacts stored in Engram memory. - OpenSpec File Convention: Specifies the openspec/ directory structure, artifact file paths per phase, delta spec sections, and archive rules. - Persistence Contract: Resolves artifact store modes (engram, openspec, hybrid, none) and defines read/write behavior, sub-agent context rules, and state persistence per mode. - Use Case: When an orchestrator launches an sdd-propose sub-agent in hybrid mode, the sub-agent reads these shared references to know exactly where to write the proposal file and how to save it to Engram with the correct topic_key. ## Quick Start Read the relevant shared reference file such as persistence-contract.md before implementing or invoking any SDD phase skill.

Frequently Asked Questions about _shared

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

FAQPage Schema
How do SDD skills persist artifacts across sessions?▼

SDD skills persist artifacts using mem_save with a deterministic topic_key like sdd/{change-name}/{artifact-type}, enabling upserts and cross-session recovery. Recovery requires mem_search to get the observation ID, then mem_get_observation for full content since search returns only truncated previews.

What is the difference between engram, openspec, and hybrid persistence modes?▼

Engram mode stores artifacts only in local memory with upsert overwrites and no history. OpenSpec mode writes files to the repo for git history and team sharing. Hybrid mode writes to both, giving cross-session recovery plus version-controllable files at higher token cost.

Can the _shared skill be invoked directly?▼

No, _shared is a support package only and is explicitly marked as not invokable with disable-model-invocation and user-invocable set to false. It exists solely to store reference documents consumed by real SDD phase skills.

Why does mem_search not return full artifact content?▼

mem_search returns only 300-character previews by design, so skills must call mem_get_observation with the observation ID to retrieve complete artifact content. Using search previews as source material produces incorrect downstream output.

What happens when re-running an SDD phase in engram mode?▼

Re-running a phase overwrites the previous artifact because Engram uses topic_key-based upserts with no revision history. For iteration history or team collaboration, the persistence contract recommends openspec or hybrid mode instead.