docs(agentos): author the coordination bigger-picture guide — how the flat-peer institution actually runs (team-of-teams)

Part of #14310 · content rubric + §Mandatory V-B-A grounding in #14310. Refs #14310.

Problem: there's no bigger-picture guide tying together HOW the institution coordinates: A2A + wakes/hooks + skills + Golden Path + cross-family review + the flat-peer / gardener model. Readers see pieces, never the whole running system — the differentiator that's hardest to grasp and most worth showing.

Acceptance Criteria

• New guide: the team-of-teams (gardener) coordination model — how A2A, wakes/hooks, the core skills, the Golden Path, and cross-family review compose into a self-running institution; positioned vs the industry's "loop engineering is the answer" (it relocates the human; a flat-peer quorum doesn't).
• The night shift as lived proof (10–20 PRs/night, no operator); honest (gated-RSI, human merge-gate by governance choice).
• Ties the coordination-mechanics cluster (A2A / hooks / skills guides) into one picture; sits high in the Brain reading order; why-care.
• Registered in tree.json.

⚠️ V-B-A grounding (mandatory before writing): the A2A / wake / hooks / skills substrate (cross-ref the sibling guides once grounded), the v13.0.0 release notes (night-shift + loop-engineering framing), AGENTS.md swarm-topology anchor (flat-peer model). Write from VERIFIED reality; capture grounding evidence in the PR.

Files: new learn/agentos/ coordination guide (name TBD) + learn/tree.json.