decision-migration

Migrate per-project manager decision records into a personal repository with dry-run verification.

6|Updated May 18, 2026
One-click install
npx skills add https://github.com/mokhtarabadi/cognitive-lead-hq --skill decision-migration-mokhtarabadi
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: decision-migration
Source: https://github.com/mokhtarabadi/cognitive-lead-hq/tree/main/skill-templates/decision-migration
Command: npx skills add https://github.com/mokhtarabadi/cognitive-lead-hq --skill decision-migration-mokhtarabadi

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Projects accumulate manager decisions in their local .opencode/decisions/ store, but the personal repository is the authoritative personality source. This Skill moves those records safely — scrubbed, verified, de-duplicated, and append-only — without risking data loss or silent corruption. ## Core Features & Use Cases - Dry-Run-First Migration: Scans every source record and classifies it as would-migrate, would-skip, or would-reject before writing anything, requiring explicit Manager batch approval. - Idempotent & Append-Only: Skips records already present in the target via migrated_from provenance checks, never edits source or target history, and proves source immutability with git status --porcelain. - Validated Write Path: Persists records only through record_manager_decision, so scrubbing, schema validation, and INDEX regeneration ride along on every write. - Use Case: When onboarding an existing project to the personal repo, invoke this Skill to consolidate all historical manager decisions into the authoritative store with a full audit trail. ## Quick Start Tell the Manager to call the migration skill for this project, then review the dry-run table and approve the batch to migrate the decision records.

Frequently Asked Questions about decision-migration

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

FAQPage Schema
How do I migrate manager decisions from a project to the personal repo?▼

Invoke the migration skill inside the project; it immediately runs a dry run classifying each record as would-migrate, would-skip, or would-reject. After you approve the batch, records are written through record_manager_decision with provenance metadata.

What happens if I run the decision migration twice?▼

The migration is idempotent. Before writing, each record is checked for an existing identical migrated_from value in the target, and matching records are skipped, so reruns change nothing.

Does the migration modify the original project decision store?▼

No. The source store is never edited; the skill proves source immutability with git status --porcelain before and after the run. Corrections on either side are new append-only tombstone records, never edits or deletes.

What happens when a decision record fails validation during migration?▼

A failed scrub or schema verification blocks only that record, which is reported with its reason in the final counts. The rest of the batch proceeds normally and no record is dropped silently.

When does the migration require a task file?▼

Runs of 50 records or fewer proceed without a task file. A backlog task file is created only when the source holds more than 50 records or the Manager explicitly requests one.