Consolidate codex-derived learnings into .beads/learnings with provenance checks.
npx skills add https://github.com/minhtri2710/skills --skill beo-dream
Please help me install this Agent Skill. Skill: beo-dream Source: https://github.com/minhtri2710/skills/tree/main/skills/beo/dream Command: npx skills add https://github.com/minhtri2710/skills --skill beo-dream
Consolidates codex-derived learnings into a stable, durable knowledge store under .beads/learnings, reducing drift and enabling governance over prior decisions.
Run a manual dream consolidation pass to refresh and lock in durable learnings across the current feature set, using the default 30-day staleness window unless overridden.
High-intent search queries and answers about installing and using this skill.
Consolidating learnings requires a manual consolidation pass that writes durable knowledge into the .beads/learnings store, applying provenance checks and governance rules to determine whether to update or create new entries.
To prevent knowledge drift, use a recurring consolidation mode with a default 30-day staleness window, ensuring prior decisions are refreshed, locked in, and governed within the durable knowledge store.
Redaction automatically scrubs secrets and PII from learnings before summaries are generated or writes are persisted, ensuring sensitive data never enters the .beads/learnings markdown files.
Editing critical patterns requires explicit approval, as the consolidation process enforces a critical-patterns approval gate to prevent unauthorized modifications to governed knowledge entries.
Bootstrap mode is needed for initial knowledge store setup, while recurring mode handles subsequent consolidation passes with a 30-day staleness window to refresh and update existing learnings.
The knowledge store protocol integrates via references/knowledge-store.md, keeping durable knowledge focused in .beads/learnings markdown files while persisting run provenance after each consolidation pass.
1012LoninLog learnings, errors, and corrections to .learnings with standardized metadata.
minhtri2710Generate structured learnings from .beads artifacts and git history.
anionzoExtract reusable patterns, decisions, and failures into a three-category knowledge document.
subroh0508Identify gaps in harness assets and generate dated improvement proposals.
