[codex] Add memory triage and inherited subagent skills

.gitignore

@@ -157,3 +157,4 @@ Thumbs.db
 node_modules/
 site/.next/
 site/out/
+.venv-ci/

CHANGELOG.md

@@ -6,6 +6,161 @@ All notable changes to this project are documented here. The product is
 
 ## Unreleased
 
+### Fixed — Repo-wide consistency audit (2026-06-18)
+
+- **Drive path unification (P0)** — removed duplicate `SwarmDriveManager` from `drive.py`; `get_default_drive()` and `create_scoped_pool()` now use `swarm_manager.get_or_create_pool()` → shared `swarms/<id>/drive/` (aligns MCP, memory bank, Experience Graph).
+- **`__init__.py`** — `SwarmDriveManager` / `get_swarm_drive_manager` imported from `swarm_manager` only.
+- **Dreaming defaults** — `ingestion.py` / `dilation.py` use `get_default_drive_path()` / `get_genomes_dir()` instead of legacy `~/.agentdrive/pool/`.
+- **Docs** — `ARCHITECTURE.md`, `CAPABILITY_FUNNEL.md`, `SWARM.md`, `FOR_AI_MODELS.md`, `rules-and-patterns.md`, `SETTINGS.md`, `ASSESSMENT.md`, `DEVELOPERS.md`, `HELP.md` updated for v2 shared Drive + full capability funnel.
+- **`docs/docs.json`** — navigation trimmed to existing pages; added Memory Bank, Capability Funnel, Skills Library groups.
+- **Tests** — `test_create_scoped_pool_uses_shared_swarm_drive` in `test_shared_swarm_drive.py`.
+- **Examples/scripts** — ruff clean across `scripts/` and `examples/`.
+
+### Fixed — CI ruff lint + format (2026-06-18)
+
+- **ruff check** — removed unused imports (`json`, `list_projects`, `ingest_from_operation`, `ingest_observation_mirror`, `PathTraversalError`); fixed import ordering across cognition, memory, learning, and test modules.
+- **ruff format** — reformatted 43 files to pass `ruff format --check`.
+- **`codebase/framework.py`** — removed unused `category` variable in `match_against_framework`.
+
+### Added — README hero wallpaper (2026-06-18)
+
+- **`assets/agentdrive-readme-hero-1920x1080.jpg`** — new 16:9 README banner: Experience Graph network + compounding funnel layers (cyan → teal → blue → violet → amber).
+- **README.md** — hero image updated to the new asset.
+
+### Memory Systems Sprint — 2026-06-18
+
+**Branch:** `codex/memory-systems-whole-update`  
+**Theme:** Turn AgentDrive from a structural memory substrate into a **full compounding intelligence platform** — memory, patterns, skills, and framework routing that grow automatically on every MCP session.
+
+This sprint completes the capability funnel:
+
+```
+Observe / Decide
+       ↓
+Experience Graph (structural memory + reasoning traces)
+       ↓
+Growth Merge (cross-surface pattern compounding)
+       ↓
+Memory Bank (deep personal knowledge databank)
+       ↓
+Skills (learned + born fused playbooks)
+       ↓
+Genomes / DNA (versioned, promotable packages)
+```
+
+#### Added — Memory Bank (deep AI knowledge databank)
+
+- **`agentdrive.memory`** — per-swarm append-only `memory_bank/memories.jsonl` with kinds: `fact`, `insight`, `decision`, `pattern`, `born_skill`, `learning`, `episode`, and more.
+- **AgentDrive-native scoping:** `vault` / `topic` / `origin_path` / `shard_index` / `preserves_source` on `MemoryEntry` — no ported metaphors, no legacy aliases.
+- **Modules:** `scope.py`, `ranking.py`, `anchor.py`, `relations.py`, `dialogue_import.py` (BM25 search, session anchor, time-bounded relations, dialogue import).
+- **Auto-ingest** from `auto_absorb`, skill fusion, and `learnings_log` (`AGENTDRIVE_AUTO_MEMORY_BANK=1` default).
+- **Ops:** `memory_bank_store`, `memory_bank_recall`, `memory_bank_search`, `memory_bank_list`, `memory_bank_briefing`, `memory_bank_deep_briefing`, `memory_bank_stats`, `memory_bank_anchor`, `memory_bank_import_dialogue`, `memory_relation_record`, `memory_relation_query`, `memory_relation_expire`.
+- **`memory_bank_deep_briefing`** — Experience Graph fabric pack + Memory Bank in one call.
+- **Docs:** `docs/MEMORY_BANK.md` (rewritten).
+- **Tests:** `tests/test_memory_bank.py`, `tests/test_memory_vault.py`.
+
+#### Added — Growth merge (experience + patterns + memory)
+
+- **`learning/growth_merge.py`** — recognizes cross-surface patterns (memory token overlap, structural similarities, codebase frameworks) and merges session growth into compound memories (`vault=growth`, `topic=merge`) plus relations.
+- **Auto-hook** in `auto_absorb` — `auto_learning.growth_merge` when ≥2 axes present (`AGENTDRIVE_AUTO_GROWTH_MERGE=1` default).
+- **Op:** `growth_merge_briefing` — unified experience + patterns + memory briefing.
+- **Tests:** `tests/test_growth_merge.py`.
+
+#### Added — Born skills (experience + skills + patterns fusion)
+
+- **`learning/skill_fusion.py`** — merges Experience Graph traces, distilled/inherited skills, and codebase pattern signals into a **new** `fused-*` skill (not a copy of any parent).
+- **Auto-fuse** in `auto_absorb` when a session spans ≥2 axes (`AGENTDRIVE_AUTO_FUSE_SKILLS=1` default); `auto_learning.fused_skill` on results.
+- **Op:** `synthesize_fused_skill(trigger, source_skills, pattern_projects, experience_traces, ...)`.
+- **Tests:** `tests/test_skill_fusion.py`.
+
+#### Added — Framework skill playbook (AgentDrive-as-framework)
+
+- **`learning/framework_skills.py`** — route `learned-*` / `fused-*` skills to tasks; unified session pack for any model using AgentDrive as its substrate.
+- **Ops:** `framework_session_start`, `framework_skill_route`, `framework_skill_run`.
+- **`when_to_call`** on auto-installed skills (from trigger/intent) so models know when to invoke each playbook.
+- **`run_skill`** accepts optional `swarm_id` for framework invocations.
+- **Docs:** `FOR_AI_MODELS.md` (Golden Rules 3–4), `SKILLS-LIBRARY.md`, `CAPABILITY_FUNNEL.md`.
+- **Tests:** `tests/test_framework_skills.py`.
+
+#### Changed — Descriptive learned skill names
+
+- **`learning/skill_naming.py`** — human-readable slugs replace opaque `auto-*` and hash-suffixed `fused-session-*` names.
+- **Learned:** `learned-{project}-{verb}-{focus}` (e.g. `learned-openmangos-mimic-growth-merge-briefing`).
+- **Fused:** `fused-{project}-{axes}` (e.g. `fused-openmangos-experience-patterns-skills`).
+- **Tests:** `tests/test_skill_naming.py`.
+
+#### Removed — Legacy memory field aliases
+
+- **`MemoryEntry.from_dict`** — no longer reads `wing` / `room` / `source_file` / `chunk_index` / `verbatim`; canonical fields only.
+- **Ops** — `memory_bank_search` / `memory_bank_anchor` / `memory_bank_import_dialogue` accept `vault` / `topic` only.
+- **Deleted:** `docs/MEMPALACE_INTEGRATION.md`, `tests/test_mempalace_integration.py`.
+
+#### Changed — Consolidation sprint (architectural audit)
+
+- **Archived** 8× `STABILIZATION_SUBAGENT_REPORT-*.md` + stale `BUILD_STATUS.md` / `MISSION_PLAN.md` → `archive/development-history/`.
+- **`docs/CAPABILITY_FUNNEL.md`** — single funnel doc with Growth Merge + Memory Bank tiers.
+- **`docs/ARCHITECTURE.md`** — subsystem map + data layout.
+- **Docs reframed** Pool→Drive terminology in `POOL.md`, `SWARM.md`, `SETTINGS.md`, `ASSESSMENT.md`.
+- **`pyproject.toml`** — `asyncio` pytest marker + `asyncio_mode = auto`.
+- **Tests:** `tests/test_multiverse_engine.py`.
+
+#### Added — Mirror-neuron codebase mimicry
+
+- **`codebase/mirrors.py`** — observation activates motor programs; cross-project `mirror_resonance.json` universal priors; Experience Graph traces on each fire.
+- **`codebase/exemplars.py`** — extract concrete motor templates from observed source.
+- **Ops:** `codebase_mimic`, `codebase_transform_style`, `codebase_mirror_resonance`.
+- **Tests:** `tests/test_mirror_neurons.py`.
+
+#### Added — Codebase pattern recognition
+
+- **`agentdrive.codebase`** — register roots, observe files, crystallize pattern frameworks, match snippets before patching.
+- **Ops:** `codebase_register_project`, `codebase_observe_file`, `codebase_patterns_profile`, `codebase_patterns_match`, `codebase_list_projects`.
+- **`agentdrive_inhabitant_read_source`** auto-observes into project `agentdrive`.
+- **Storage:** `~/.agentdrive/codebase-patterns/<project>/`.
+- **Tests:** `tests/test_codebase_patterns.py`.
+
+#### Added — End-to-end automatic learning
+
+- **`agentdrive.learning.auto_absorb`** — post-`run_operation` hook (`AGENTDRIVE_AUTO_LEARN=1` default): session tracking, auto reasoning traces, Hermes-style skill distillation, promote + DNA ingest on high-signal ops.
+- **Results** include `auto_learning` summarizing absorbed traces, skills, genomes, growth merge, and fused skills.
+- **Tests:** `tests/test_auto_learning.py`.
+
+#### Added — Multiverse cognition + external MCP Parent
+
+- **`src/agentdrive/cognition/`** — `MultiverseEngine`, `MultiverseSessionStore`, 7-phase pipeline (spawn → simulate → invariants → stress-test → collapse → record).
+- **External Parent** — connected MCP models (Grok, Claude, Cursor) submit branch reasoning via `external_parent_decision`; persisted as `llm_mode=external`.
+- **Local LLM spawner** — `~/.agentdrive/local_models.yaml` for branch simulation; heuristic fallback.
+- **Ops:** `multiverse_parent_decision`, `multiverse_run_full`, `multiverse_list_sessions`, `multiverse_get_session`, `multiverse_densify`, `multiverse_reopen_stale`, `external_parent_decision`.
+- **CLI:** `agentdrive multiverse run|list|status`
+- **Docs:** `docs/MULTIVERSE_COGNITION.md`; **example:** `examples/12_multiverse_cognition_loop.py`, `examples/14_external_mcp_parent_loop.py`.
+- **Tests:** `tests/test_external_parent_decision.py`, `tests/test_multiverse_engine.py`.
+
+#### Fixed — Handler `swarm_id` duplicate kwarg
+
+- `memory_bank_anchor`, `growth_merge_briefing`, `framework_session_start` — pop `swarm_id` from pack before `_success()` to avoid duplicate keyword argument.
+
+#### Verified — OpenMangos integration (post-restart)
+
+- **Swarm:** `mangos-pablothethinker-openmangos` · **Project:** `openmangos`
+- **11/11 ops passed** including framework ops (`framework_session_start`, `framework_skill_route`, `framework_skill_run`) plus memory/growth ops.
+- **18 learned skills** on bench; top route matches: `learned-openmangos-mimic-wire-agentdrive-growth-merge-a`, `fused-openmangos-experience-patterns-skills`.
+
+#### Verify (local)
+
+```bash
+cd "Vektra Industries/Software/AgentDrive"
+PYTHONPATH=src python -m pytest tests/test_memory_bank.py tests/test_memory_vault.py \
+  tests/test_growth_merge.py tests/test_skill_fusion.py tests/test_skill_naming.py \
+  tests/test_framework_skills.py tests/test_auto_learning.py tests/test_mirror_neurons.py \
+  tests/test_codebase_patterns.py tests/test_external_parent_decision.py tests/test_multiverse_engine.py -q
+PYTHONPATH=src python examples/12_multiverse_cognition_loop.py --trigger "Ship multiverse MVP"
+PYTHONPATH=src python -m agentdrive.cli multiverse run --trigger "CLI test" --branches 5
+```
+
+**Env toggles (all default on):** `AGENTDRIVE_AUTO_LEARN=1`, `AGENTDRIVE_AUTO_MEMORY_BANK=1`, `AGENTDRIVE_AUTO_GROWTH_MERGE=1`, `AGENTDRIVE_AUTO_FUSE_SKILLS=1`
+
+---
+
 ### Fixed — `scripts/install.sh` piped install (`curl | bash`)
 
 - **`scripts/install.sh`** — when piped (no `BASH_SOURCE`), fetches canonical `install.sh` from GitHub instead of failing with `BASH_SOURCE[0]: unbound variable` and `//install.sh` path error.

DEVELOPERS.md

@@ -1,6 +1,6 @@
 # DEVELOPERS — One-Command Bring-Up
 
-This is the contributor's quick-start. For high-level orientation read [`VISION.md`](VISION.md), then [`ARCHITECTURE.md`](ARCHITECTURE.md). For the project's public framing read [`README.md`](README.md).
+This is the contributor's quick-start. For high-level orientation read [`VISION.md`](VISION.md), then [`docs/ARCHITECTURE.md`](docs/ARCHITECTURE.md) and [`docs/CAPABILITY_FUNNEL.md`](docs/CAPABILITY_FUNNEL.md). For the project's public framing read [`README.md`](README.md).
 
 ## Prerequisites
 

HELP.md

@@ -514,7 +514,7 @@ These are production optional advanced extensions. Core AgentDrive (three drives
 
 **Deep reference & mental models**:
 - `CONCEPTS.md` (the single best deep, agent-and-human-friendly explanation of the three-tier model, genomes, living loops, quarantine, reconciliation, and the new Lineage-enhanced components)
-- `ARCHITECTURE.md`
+- `docs/ARCHITECTURE.md`
 - `GENOME-SPEC.md`
 - `docs/AGENTDRIVE-V2-INHERITANCE.md`
 - `docs/SETTINGS.md`