v2.0.0 — Never explore twice - #29

Merged

RodCor merged 27 commits into

mainfrom

release/v2.0.0

Jun 22, 2026

Commits on Jun 20, 2026

chore(release): open v2.0.0 — bump workspace version 1.5.1 -> 2.0.0

RodrigoCordoba
and
claude
committed
ci(release): S4.7 — version-guard + binary-version assert + Node-24 actions + bump helper

RodrigoCordoba
and
claude
committed
docs(hosts): S4.5 — verify Cursor/Gemini MCP schemas against live docs, drop caveat

RodrigoCordoba
and
claude
committed
feat(s4): four hardening minors — superseded filters, atomic sidecars, toml_edit config writes

RodrigoCordoba
and
claude
committed
merge S4.5 + S4.7 (host verification + release guards) into release/v2.0.0
RodrigoCordoba
committed
merge S4.1-S4.4 (hardening minors) into release/v2.0.0
RodrigoCordoba
committed
feat(S1): local-model independence — Ollama provider, [cheap_model] section, doctor probe

RodrigoCordoba
and
claude
committed
merge S1 (local-model independence: Ollama provider + [cheap_model] config + doctor/docs) into release/v2.0.0
RodrigoCordoba
committed
feat(retrieval): S5.1 — RetrievalBackend trait + FlatBackend + StorageSection config seam

RodrigoCordoba
and
claude
committed
feat(retrieval): S5.2 — GraphLiteBackend + memory_edges typed-edge projection (schema v4)

RodrigoCordoba
and
claude
committed
merge S5.1+S5.2 (RetrievalBackend trait + graph-lite Tier 1 backend, schema v4) into release/v2.0.0
RodrigoCordoba
committed
feat(flagship1): Pass A — episodic work-resume core (schema v5)

RodrigoCordoba
and
claude
committed
feat(flagship1): Pass B — repo digest + SessionStart hook + ROI attribution

RodrigoCordoba
and
claude
committed
merge Flagship 1 (session warm-start: episodic resume + repo digest + SessionStart hook, schema v5) into release/v2.0.0
RodrigoCordoba
committed
feat(brain): Flagship 3 Pass A — kimetsu ask + brain-answer MCP tool + command fast-path

RodrigoCordoba
and
claude
committed

Commits on Jun 21, 2026

Commits on Jun 22, 2026

docs(release): v2.0 README + config reference — lead with metrics & configuration

RodrigoCordoba
and
claude
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

v2.0.0 — Never explore twice - #29

v2.0.0 — Never explore twice#29
RodCor merged 27 commits into
mainfrom
release/v2.0.0

Commits on Jun 20, 2026

chore(release): open v2.0.0 — bump workspace version 1.5.1 -> 2.0.0

ci(release): S4.7 — version-guard + binary-version assert + Node-24 actions + bump helper

docs(hosts): S4.5 — verify Cursor/Gemini MCP schemas against live docs, drop caveat

feat(s4): four hardening minors — superseded filters, atomic sidecars, toml_edit config writes

merge S4.5 + S4.7 (host verification + release guards) into release/v2.0.0

merge S4.1-S4.4 (hardening minors) into release/v2.0.0

feat(S1): local-model independence — Ollama provider, [cheap_model] section, doctor probe

merge S1 (local-model independence: Ollama provider + [cheap_model] config + doctor/docs) into release/v2.0.0

feat(retrieval): S5.1 — RetrievalBackend trait + FlatBackend + StorageSection config seam

feat(retrieval): S5.2 — GraphLiteBackend + memory_edges typed-edge projection (schema v4)

merge S5.1+S5.2 (RetrievalBackend trait + graph-lite Tier 1 backend, schema v4) into release/v2.0.0

feat(flagship1): Pass A — episodic work-resume core (schema v5)

feat(flagship1): Pass B — repo digest + SessionStart hook + ROI attribution

merge Flagship 1 (session warm-start: episodic resume + repo digest + SessionStart hook, schema v5) into release/v2.0.0

feat(brain): Flagship 3 Pass A — kimetsu ask + brain-answer MCP tool + command fast-path

Commits on Jun 21, 2026

feat(brain): Flagship 3 Pass B — answer-grade injection (3.3) + proactive pre-fetch toggle (3.5)

merge Flagship 3 (the active brain: ask + brain_answer MCP tool + command fast-path + answer-grade injection + opt-in pre-fetch) into release/v2.0.0

feat(brain): Flagship 2 — Memory → Skill synthesis (2.1–2.5)

merge Flagship 2 (memory to skill synthesis: trigger + draft + review/install + staleness, schema v6) into release/v2.0.0

feat(brain): Epic S2 — Continuous self-tuning + ROI v2

merge S2 (continuous self-tuning: re-tune triggers + model advisor + regret-penalty objective + ROI v2) into release/v2.0.0

feat(sync): Epic S3 — personal brain sync via event-log replication

merge S3 (personal brain sync: event-log replication export/import + directory protocol) into release/v2.0.0

feat(backend): S5.3 PetgraphBackend (graph feature, remote-only) + S5.4 cross-backend benchmark

merge S5 remote-track (PetgraphBackend Tier-2 feature-gated + cross-backend benchmark harness) into release/v2.0.0

docs(changelog): v2.0.0 — Never explore twice

Commits on Jun 22, 2026

docs(release): v2.0 README + config reference — lead with metrics & configuration