mempalace-code 1.14.2__tar.gz → 1.16.0__tar.gz

{mempalace_code-1.14.2 → mempalace_code-1.16.0}/.gitignore

@@ -4,6 +4,8 @@ build/
 __pycache__/
 *.pyc
 .pytest_cache/
+.mypy_cache/
+.ruff_cache/
 mempal.yaml
 mempalace.yaml
 entities.json
@@ -25,6 +27,14 @@ venv/
 .env
 .verify-state
 
+# Local agent tooling state: absolute local paths and private notes
+.codex/
+.codex-local/
+.playwright-mcp/
+
+# Default `dependency_upgrade_gate.py current-audit` output (a workflow artifact)
+dependency-audit-output/
+
 # IDE
 .idea/
 .vscode/
@@ -49,7 +59,7 @@ autopilot.toml
 benchmarks/results_token_delta_*.json
 !benchmarks/results_token_delta_mempalace.json
 
-# Claude Code session data (keep skills, prompts, settings.json tracked)
+# Claude Code session data (keep skills and prompts tracked; a project settings.json may be tracked)
 .claude/*
 !.claude/skills/
 !.claude/prompts/

{mempalace_code-1.14.2 → mempalace_code-1.16.0}/.pre-commit-config.yaml

@@ -2,7 +2,7 @@ repos:
   - repo: https://github.com/astral-sh/ruff-pre-commit
     rev: v0.15.16
     hooks:
-      - id: ruff
+      - id: ruff-check
         args: [--fix]
         files: ^(mempalace_code|tests|scripts)/.*\.py$
       - id: ruff-format

mempalace_code-1.16.0/AGENTS.md

@@ -0,0 +1,304 @@
+# mempalace — Agent Guide
+
+## 0. Rule Zero — smallest justified solution
+
+Rule Zero is owned by the "Rule Zero — smallest justified solution" section of the
+maintainer's user-level `~/.codex/AGENTS.md`; Claude Code reads the identical
+section from `~/.claude/CLAUDE.md`. It is not restated here; every rule in this file
+is a local addition to it.
+
+**Release credential boundary.** Release preparation, qualification, admission,
+and publication gates must never execute external AI clients such as `codex`,
+`claude`, or `gemini`; run their authentication commands or provider/model calls;
+or read, copy, inspect, require, or transmit user credentials, API keys, OAuth
+tokens, keychains, or paid-account state. Verify MemPalace with its local CLI,
+credential-free stdio MCP protocol checks, installed-package behavior, and
+credential-free hosted workflows. An optional interoperability exercise with an
+AI client requires separate explicit owner authorization, runs outside the
+release gate, and cannot block a release.
+
+**Direct functional release boundary.** Before release candidate publication,
+run the exact installed wheel through its real CLI and credential-free stdio MCP
+interfaces in disposable environments. Reconcile every discovered command and
+subcommand, every enabled tool in every profile, and supported runtime extras
+with concrete success or documented refusal evidence and relevant failure/retry
+postconditions. Help, parser errors, tool listings, handler-only calls and source
+pytest counts do not prove functional execution. Missing required coverage blocks
+release even when CI passes. Record confirmed problems through backlog-utility
+and retain a public-safe acceptance report. Follow `docs/RELEASING.md` section 1a.
+
+**Commit attribution boundary.** Commits in this repository must not contain a
+Claude `Co-Authored-By` trailer. Configure automation to omit it; do not add it
+and clean it up later.
+
+## Audience and default scope
+
+- The default audience is developers asking questions about repository code.
+  Ground answers in current source, tests, repository documentation, and Git history.
+- Users preserve valuable project results in repository artifacts and commits.
+  Use those version-controlled artifacts as durable project context.
+- Inspecting or mining stored user conversations, maintaining diaries, and adding,
+  enabling, or repairing conversation-save reminders or persistence guarantees
+  require an explicit request for that workflow. An audit finding about an optional
+  conversation workflow does not bring it into a code-question task.
+
+## Stack
+
+- **Python** 3.11+ (supports 3.11–3.14)
+- **Storage**: LanceDB (core, crash-safe vector DB — no server required)
+- **Embeddings**: FastEmbed/ONNX by default; SentenceTransformer only in `[custom-models]`
+- **MCP**: the official `mcp` Python SDK
+- **Config**: PyYAML
+- **Runtime dependencies**: declared once in `[project].dependencies` of `pyproject.toml`
+- **Linting / formatting / typing**: Ruff + Pyright
+- **Tests**: pytest
+- **Package manager**: uv (preferred) or pip
+
+## Dev Setup
+
+```bash
+# With uv (preferred)
+uv pip install -e ".[dev,spellcheck,treesitter]"
+
+# With pip
+pip install -e ".[dev,spellcheck,treesitter]"
+```
+
+These are the extras the CI type check installs; with `.[dev]` alone Pyright
+reports the Tree-sitter imports as missing.
+
+No Docker required. Everything runs locally in a venv or with pipx.
+
+Optional extras:
+
+- `.[custom-models]` — arbitrary SentenceTransformer models; follow `docs/OFFLINE_USAGE.md`
+- `.[dev]` — test, lint, type-check, build, and release tooling
+- `.[spellcheck]` — autocorrect helper API (`mempalace_code.spellcheck`); mining never applies it
+- `.[treesitter]` — Tree-sitter AST parsing
+- `.[watch]` — automatic mining on file changes
+
+## Running Tests
+
+**Important**: Use the Python from your mempalace virtualenv (pipx venv, `.venv`, etc.), not the system Python which may lack dev deps.
+
+```bash
+# Full suite (stop on first failure)
+python -m pytest tests/ -x -q -m "not needs_network"
+
+# Per-module
+python -m pytest tests/test_storage.py -v
+python -m pytest tests/test_mcp_server.py -v
+python -m pytest tests/test_miner.py -v
+python -m pytest tests/test_convo_miner.py -v
+```
+
+## Linting / Formatting
+
+```bash
+# Check
+ruff check mempalace_code/ tests/ scripts/
+
+# Format check
+ruff format --check mempalace_code/ tests/ scripts/
+
+# Type check (gating in CI — must exit 0)
+python -m pyright --pythonpath "$(python -c 'import sys; print(sys.executable)')"
+python -m pyright -p pyrightconfig.strict.json
+
+# Auto-fix lint
+ruff check --fix mempalace_code/ tests/ scripts/
+
+# Auto-fix format
+ruff format mempalace_code/ tests/ scripts/
+```
+
+Line length: 100. Target: py311. Quote style: double.
+
+The complete gate set the **Tests** workflow runs on every pull request is listed
+once, in `CONTRIBUTING.md` under "Running Tests".
+
+## Key Modules
+
+The main owners under `mempalace_code/`. The full module guide is
+`mempalace_code/README.md`.
+
+| Module | Purpose |
+|--------|---------|
+| `storage.py` | LanceDB vector storage — add, search, delete, health_check, recover |
+| `backup.py` | Tarball backup/restore — `mempalace-code backup`, scheduled backups |
+| `mining/` | Project miner package — scanning, language detection, chunking, symbols, batching, orchestration; `miner.py` is a compatibility shim |
+| `convo_miner.py` | Conversation miner — ingests Claude/ChatGPT/Slack exports |
+| `searcher.py` | Semantic search — query palace with optional wing/room filters |
+| `knowledge_graph.py` | Temporal KG — entity-relationship triples with validity windows |
+| `layers.py` | Tiered context loading — L0/L1/L2/L3 wake-up layers for local models |
+| `palace_graph.py` | Graph traversal and tunnel detection across wings/rooms |
+| `mcp/` | MCP server package — `registry.py` (the `TOOLS` dict), `dispatch.py`, `runtime.py`, `tools/`; `mcp_server.py` is the public entrypoint shim and `mcp_launcher.py` backs `mempalace-code-mcp` |
+| `watcher.py` | File watcher — `watch_and_mine`, `watch_all`, launchd/cron schedule rendering |
+| `cli_commands/` | CLI command handlers; `cli.py` is the thin `mempalace-code` entry point and compatibility facade |
+
+## Architecture Principles
+
+1. **Verbatim-first** — store content exactly as provided; do not summarize or compress drawers.
+2. **Local-first** — no external APIs required; default embeddings run on-device via CPU FastEmbed/ONNX.
+3. **Zero-API-by-default** — LanceDB and FastEmbed work offline after explicit model setup.
+
+## Storage Backend
+
+- **LanceDB** is the core backend (installed by default via `lancedb>=0.20`).
+- **ChromaDB** support is retired from current packages because every available
+  release is advisory-affected. Back up a legacy source palace before upgrading,
+  then use the last public bridge release in isolation:
+  `uvx --python 3.12 --from 'mempalace-code[chroma]==1.13.4' mempalace-code migrate-storage SRC DST --verify`.
+
+## Embedding Model Policy
+
+- **Current default**: `all-MiniLM-L6-v2` (384d, normalized CPU FastEmbed/ONNX).
+- **Cache authority**: `$HF_HOME/mempalace-fastembed/all-MiniLM-L6-v2-v1/` plus its
+  immutable `.mempalace-model.json` provenance. Recovery is
+  `mempalace-code fetch-model` while online.
+- **Custom-model boundary**: arbitrary Hugging Face names and local
+  SentenceTransformer paths use the explicit `[custom-models]` extra. On CPU-only Linux,
+  follow the ordered installation and recovery contour in `docs/OFFLINE_USAGE.md`; only
+  that explicit custom-model path may use `trust_remote_code=True`, and canonical MiniLM
+  aliases never do.
+- **No code-only models**: CodeBERT, UniXcoder, etc. improve code at the expense of prose. Only general-purpose sentence-transformers that handle both are candidates.
+
+### Text gate
+
+An embedding model change is gated behind A/B benchmark results
+(`benchmarks/embed_ab_bench.py`) and must pass both the code gate (the A/B
+code-retrieval results) and the text gate: match or beat MiniLM on LongMemEval R@5
+(text retrieval). Text quality is non-negotiable — this is a code-first fork but
+natural language search (conversations, commits, decisions) must not degrade. The
+only LongMemEval harness, `benchmarks/longmemeval_bench.py`, is the legacy upstream
+ChromaDB-based script. It imports `chromadb`, which current packages do not install
+and whose available releases carry known advisories, so it runs only in an isolated,
+disposable legacy environment. No text baseline has been recorded, so a model change
+must first record one there. The gate cannot be skipped: without that environment and
+baseline it has not run, and the model change is blocked.
+
+### Historical Benchmark Results — 2026-04-09 (BENCH-EMBED-AB)
+
+Historical record, measured with the fixed 20-query set that
+`benchmarks/embed_ab_bench.py` used at the time. It is not comparable with the current
+`benchmarks/code_retrieval_bench.py` dataset (symbol and file-only truth,
+schema-2 report), and no current baseline has been recorded.
+Record a fresh dated baseline on the current dataset before the next model
+decision.
+
+Code retrieval on the mempalace repo (20 known-answer queries, 469 chunks):
+
+| Model              | R@5   | R@10  | Embed(s) | Query(ms) | Index(MB) |
+|--------------------|-------|-------|----------|-----------|-----------|
+| all-MiniLM-L6-v2   | 0.950 | 1.000 | 15.2     | 15.9      | 17.0      |
+| all-mpnet-base-v2  | 0.900 | 1.000 | 47.5     | 30.5      | 17.7      |
+| nomic-embed-text-v1.5 | 0.950 | 1.000 | 85.4  | 45.8      | 17.7      |
+
+Per-category R@5:
+
+| Model              | architecture | class_lookup | cross_file | function_lookup |
+|--------------------|:---:|:---:|:---:|:---:|
+| all-MiniLM-L6-v2   | 0.800 | 1.000 | 1.000 | 1.000 |
+| all-mpnet-base-v2  | 0.600 | 1.000 | 1.000 | 1.000 |
+| nomic-embed-text-v1.5 | 1.000 | 1.000 | 1.000 | 0.833 |
+
+**Recommendation: minilm remains default.**
+
+- mpnet regresses on code R@5 (0.900 vs 0.950) while being 3× slower to embed and 2× slower at query. Eliminated.
+- nomic ties minilm on code R@5 (0.950) but is 5.6× slower to embed and 2.9× slower at query, with a 550 MB model vs 80 MB. No net gain.
+- [Text-gate](#text-gate) (LongMemEval) evidence was not collected — prerequisites missing (`benchmarks/data/longmemeval_s_cleaned.json` not present, `fastembed` not installed).
+- Full results: `benchmarks/results_embed_ab_2026-04-09.json`.
+
+## Git Workflow
+
+- **Two histories**: local `main` is the development trunk; work reaches it from
+  topic branches. Public `main` carries one squashed commit per release and advances
+  only by fast-forwarding a release candidate built and admitted per
+  `docs/RELEASING.md`.
+- **Branch naming**: `feat/<slug>`, `fix/<slug>`, `chore/<slug>`, `docs/<slug>` for
+  development; `release/vX.Y.Z` (rebuilds `release/vX.Y.Z-rc2`, `-rc3`, …) for
+  public release candidates only.
+- **Conventional commits**:
+  - `feat:` — new feature or capability
+  - `fix:` — bug fix
+  - `docs:` — documentation only
+  - `test:` — test additions or changes
+  - `bench:` — benchmarks
+  - `chore:` — maintenance, deps, tooling
+  - Release candidate commits use the subject `release vX.Y.Z`.
+- **No force-push to `main`**.
+- Pull requests on GitHub, including Dependabot updates, are never merged on
+  public `main`: the next release candidate would silently revert them. Port an
+  accepted change into the development trunk, ship it in the next release, then
+  close the pull request with a note naming that release.
+
+## Maintainer-Local Tools
+
+The `autopilot` and `backlog-utility` tools are maintainer-local; they are not
+distributed with this repository. `/task-plan` and `/task-hardening` admit
+executor state only through `autopilot` and stop when it is unavailable.
+Current task contracts and planning are owned by `.backlog` (backlog-files/v2).
+Read `backlog_context`, `backlog_workset_context`, and `backlog_workset_queue`
+through the verified project-bound MCP connection. CLI fallback uses a verified
+absolute `backlog-utility` binary with `call --store .backlog --project
+mempalace-code --principal <principal> --name <tool> --arguments-file <json>`.
+Unavailable or mismatched access means unknown backlog state; never fall back
+to active YAML or edit canonical JSON by hand. `docs/BACKLOG.yaml` retains section
+metadata; `docs/BACKLOG-archived.yaml` retains historical completion evidence.
+Recovery: `<absolute-backlog-utility> validate --store .backlog`.
+Edits still require admitted authority and no conflicting writer.
+
+## Operational Lessons
+
+- **Keep this file and public docs public-safe.** `AGENTS.md` is the canonical public instruction file; `CLAUDE.md` is only its pointer. Public docs may name package ranges, workflow categories, advisory IDs, and reproducible commands. Private remotes, hostnames, tokens, credentials, local machine paths, customer/project details, incident specifics, and non-public operational history never go here or in other public docs; put private or machine-local lessons only in an ignored local-only note outside the published tree, such as `.codex-local/LESSONS.md`.
+- **Audit privacy before every public delivery.** Follow the privacy audit in
+  `docs/RELEASING.md` section 2. Inspect the exact tracked/staged/committed tree,
+  including hidden LLM instructions, prompts, settings, examples and sealed backlog
+  history, then both distribution archives and the history reachable from the
+  outgoing ref. Secret scanning alone cannot establish that operational or personal
+  information is public-safe. Retain redacted findings and the reviewed identities;
+  a confirmed private disclosure or consequential unresolved classification blocks
+  its public delivery. Preserve private local evidence and sealed records; never
+  rewrite history, scrub canonical JSON or weaken detectors to pass this gate.
+- **Record reusable lessons only when they are public knowledge.** When a session exposes a project gotcha, publish step, verification boundary, or agent-behavior correction that is safe for public readers and useful to future contributors, rewrite the matching lesson here in place, or add one when none matches.
+- **Verify on the surface that will run the change, and name what was not run.** Python tests alone do not prove GitHub Actions runtime changes: use local YAML/static checks such as `actionlint`, then verify the real hosted workflow run when action runtime behavior matters. Direct handler calls are useful for MCP compatibility but do not prove a separate stdio MCP client, and CLI help does not prove the command executes. A release-readiness summary must distinguish unit tests, focused integration tests, direct API smoke, real CLI execution, and hosted/daemon behavior that was not run; a tag-only or release-only workflow without a real trigger run is reported as syntax-checked and version-checked, not execution-tested. Never imply full local coverage for hosted-only behavior.
+- **Check the intended public release target.** Before publishing, verify the repository, branch, tag, and workflow that public users will see. Do not assume local remote names or private mirrors represent public release truth.
+- **Treat release status as multiple independent facts.** Branch Tests, tag-triggered PyPI publish, GitHub Release creation, PyPI version visibility, and any deployment/release-environment status can diverge. Check all of them before calling a release published or latest; if one is red or missing, either fix it now or record the explicit remaining blocker.
+- **Test dependency drift with a fresh resolver.** Local `.venv` and `uv.lock` success can hide what GitHub Actions or users get from an unlocked `pip install`. For dependency-sensitive failures, reproduce in a clean pip environment matching the hosted workflow before declaring the tests fixed.
+- **Audit dependency targets before raising bounds.** For runtime, dev, and optional extras, check current and target versions against OSV or an equivalent advisory source, then run a resolver-level audit on a fresh environment. Do not raise optional legacy backends into advisory-affected ranges; hold or cap them and backlog the upgrade gate instead.
+- **Keep benchmark gates tied to measured baselines.** If a release benchmark fails, reproduce it locally against the pinned fixture, update the CI threshold only to the observed stable baseline, and backlog any desired quality increase separately.
+- **Do not call tests "local feature testing."** When asked to test new features locally, run the public CLI/MCP/API behavior itself, not only pytest. For each new feature, exercise at least one success path and one important failure/guard path when safe, record the exact command or request, and name any behavior that was covered only by tests.
+- **Clean up smoke-test artifacts immediately.** Real backup, cleanup, and benchmark smokes can create archives, temp palaces, and result JSON. Put them under disposable temp paths, verify success/failure, then remove artifacts or explicitly report what remains.
+- **Treat palace disk growth as storage forensics first.** Compare backup size, live storage stats, row counts, and cleanup output before deleting anything. Preserve non-regenerable manual drawers and KG data, stop active writers when needed, use supported cleanup APIs, and verify health/status afterward.
+
+## LDR adoption and problem handoff
+
+- LDR (`ldr/v1`) is the default format for current durable architectural decisions.
+  Canonical accepted records live under `docs/decisions` when present. Historical
+  ADRs remain evidence; do not migrate them or invent records for routine tasks.
+  Reuse unchanged applicable accepted decisions without a new proposal or repeated
+  record approval. No applicable records means an explicit empty selection with
+  its applicability reason in the existing task context.
+
+- Apply this rule before architecture, durable-contract, lifecycle, ownership or
+  recovery changes. Start with one decision needed by the current task; keep task
+  state and execution evidence with their existing owners.
+- Discover applicability from the task and repository-owned references. Resolve
+  the exact accepted key/revision; inspect status first and apply only active
+  records. Check current committed bytes, owner, local references and fitness.
+  Search results and historical prose cannot accept a decision.
+- A missing or stale binding blocks only the dependent change. Propose a record
+  outside the canonical accepted root; only the operator accepts active records.
+  Same-key changes advance revision by one and preserve lifecycle/recovery.
+- Use the existing adopter and pinned validator when present. Do not replace a
+  pin, install a utility, add a runtime gate, or migrate historical decisions as
+  part of this instruction. Run the current task's existing relevant checks.
+- For a reproduced LDR problem, prepare and deliver the targeted handoff below
+  during the same work session without waiting for a separate request. Identify
+  whether the failure belongs to LDR or the adopter; record uncertainty. Continue
+  independent safe work while the affected action remains blocked.
+
+- Private maintainer routing and the first scoped adoption candidate are in
+  .codex-local/LDR.md when present. Read that note before sending an LDR finding.
+  If no route is configured, retain the sanitized draft and report missing_route.
+  Keep machine paths, private transport and incident evidence out of public docs.

{mempalace_code-1.14.2 → mempalace_code-1.16.0}/CHANGELOG.md

@@ -2,6 +2,187 @@
 
 ## Unreleased
 
+## v1.16.0 — 2026-10-05
+
+### Added
+
+- Configure `backup_dir` independently of the palace location, with the existing
+  environment override mechanism. Manual, scheduled, pre-watch and pre-optimize
+  backups share the selected managed directory for creation, listing and rotation.
+  Each palace has an isolated directory under the selected root.
+
+### Fixed
+
+- Measure managed backup storage in the selected directory and check available
+  space on the archive destination. Invalid or inaccessible configured paths fail
+  without falling back. Existing archives stay in place, and explicit `--out`
+  keeps its rotation bypass and priority.
+- Release acceptance now executes every enabled MCP tool in every profile, checks
+  both protocol handshakes, and discovers all delegated migration actions. Help
+  and parser guidance do not count as functional execution.
+- Release socket fixtures use a short owned path on platforms with limited Unix
+  socket path lengths, with checked cleanup and explicit unsupported-platform handling.
+
+## v1.15.1 — 2026-10-04
+
+Includes the v1.15.0 changes and the corrections below.
+
+### Security
+
+- Refresh the transitive lock pins to PyJWT 2.15.0 and urllib3 2.8.0, closing
+  published authentication, TLS and denial-of-service advisories. The complete
+  current audit checks all 152 external locked packages without exceptions.
+
+### Fixed
+
+- Govern the synthetic Kubernetes Secret fixture in immutable release-candidate
+  history with one exact Gitleaks fingerprint so full-history release scanning
+  can retain that history and continue detecting other secrets.
+
+## v1.15.0 — 2026-09-26
+
+Maintenance release from a full repository and documentation audit. Search scores
+now use the true cosine scale, every surface uses the selected palace's knowledge
+graph, and several data-loss, hook, and documentation defects are fixed.
+
+### Upgrade notes
+
+- Search scores change scale; result order does not. An old score `s` corresponds
+  to `(s + 1) / 2`. Recalibrate any script or prompt that filters on a fixed score.
+- CLI commands run without `--palace` now read and write the configured palace's
+  `<palace>/knowledge_graph.sqlite3`. Facts that only the legacy global
+  `~/.mempalace/knowledge_graph.sqlite3` holds are copied in once (see Changed).
+  If you plan to rebuild a palace from a JSONL export, take that export after
+  upgrading.
+
+### Security
+
+- Raised the optional `[custom-models]` minimum to sentence-transformers 5.6.0
+  to exclude CVE-2026-68770. Default FastEmbed installations are unchanged.
+- Refreshed `uv.lock` past published advisories: anyio 4.15.1, httpx2/httpcore2
+  2.13.0, starlette 1.6.0 and click 8.5.0 (runtime, via `mcp`), and transformers
+  5.15.1, torch 2.14.0 and setuptools 84.0.0 (`[custom-models]`), with a passing
+  dependency-upgrade gate report.
+- The scheduled Dependency Audit now checks every package pinned in `uv.lock`,
+  direct and transitive, against OSV, and fails closed when an OSV or PyPI
+  yanked-status lookup is incomplete. A transitive advisory can be accepted only
+  for its exact locked version.
+- The Gitleaks gate is built with Go 1.27.1 instead of end-of-life Go 1.24.11.
+- Release builds install the reviewed `uv.lock` versions of `build` and `twine`.
+- Dependabot waits seven days before proposing version updates, and the CI zizmor
+  gate also audits `.github/dependabot.yml`.
+- Added `SECURITY.md`: report vulnerabilities through GitHub private vulnerability
+  reporting.
+
+### Changed
+
+- `mempalace_search` accepts `max_results` as an alias for `limit`. The default
+  remains five results; both values must agree when supplied together.
+- Search similarity is now true cosine similarity (`1 - distance / 2`; 1.0 =
+  identical, about 0 = unrelated) in CLI `search`, `search_memories()`,
+  `code_search()`, MCP `mempalace_search` / `mempalace_code_search`, Layer 3
+  wake-up search, and `mempalace_check_duplicate`. Up to 1.14.2 these reported
+  `1 - distance` (`2 × cosine - 1`, often negative). `code_search`'s
+  `ranking.vector_distance` still reports the raw squared-L2 distance.
+- Facts held only by the legacy global KG are copied into the configured palace's
+  KG once per legacy file state, recorded by
+  `~/.mempalace/knowledge_graph.sqlite3.adopted`. Facts the palace KG already
+  holds stay authoritative, the legacy database is never modified, other palaces
+  never receive its facts, and adoption does not block concurrent KG writers.
+  Delete the marker to adopt again.
+- `restore FILE` without `--palace` writes the archived KG to the configured
+  palace's `<palace>/knowledge_graph.sqlite3`; `--kg-path` still overrides it.
+- The package's Upstream URL now points at https://github.com/MemPalace/mempalace.
+
+### Fixed
+
+- Invalid metadata filters keep the backend query error with current LanceDB
+  releases; they no longer report a healthy palace as damaged.
+- Kubernetes Secret YAML with whitespace before a key's colon is excluded from
+  mining.
+- MCP mining creates or adopts the knowledge graph only after acquiring the mine
+  lease and when a graph write is needed. Busy and unchanged requests preserve it.
+- File context reports complete chunk totals and pagination beyond 10,000 chunks.
+- MCP storage-open failures report health/repair recovery, and a cached knowledge
+  graph reports a missing palace through the structured palace error.
+- Installation maintenance blocks palace writes from other threads while preserving
+  same-thread nesting and the supported migration fence.
+- Backup publication preserves an archive created by a concurrent writer. Platforms
+  without hard-link publication report a safe failure instead of replacing it.
+- Updater detection preserves pipx installations through aliased custom homes.
+  Installed-package checks retain the selected manager and validate recovery
+  commands against the installed launcher.
+- `import` no longer silently drops distinct drawers. Its dedup squared an
+  already-squared distance (skipping anything above about cosine 0.78), compared
+  across every wing, and let records of one file dedup each other. It now skips a
+  record only when its id is already stored or repeats in the file, or when its
+  target wing already held a drawer with cosine ≥ 0.9 before the import.
+- `mempalace_check_duplicate` and the `mempalace_add_drawer` guard now apply the
+  documented cosine ≥ 0.9 threshold; they previously matched only at about 0.95.
+- The knowledge graph no longer splits between MCP and the CLI. `mine`,
+  `mine-all`, `mine --watch`, `export --with-kg`, `import`, `backup create`,
+  `restore`, `compress`, pre-optimize and pre-watch backups, MCP, and the watcher
+  all use the selected palace's KG. Since v1.13.5, MCP could not see CLI-mined
+  facts, and default backups and exports omitted MCP-written facts.
+- `mine --full` (and MCP `mempalace_mine` with `full: true`) now removes drawers,
+  tiny-file entries, and KG facts of deleted or newly excluded files, as a plain
+  full-walk mine already did.
+- The legacy Claude Code PreCompact hook no longer blocks compaction; it logs and
+  exits without a decision. The Save hook counts only human prompts toward its
+  interval, and both hooks mine `MEMPAL_DIR` with `--mode convos`.
+- The documented rebuild validation uses `search --results 5`; `--limit` does not
+  exist, so the step always failed after the palace had been quarantined.
+- The legacy ChromaDB bridge command pins Python 3.12
+  (`uvx --python 3.12 --from 'mempalace-code[chroma]==1.13.4' ...`), avoiding a
+  chroma-hnswlib source build on Python 3.13+.
+- `mempalace-code wing-migration <action> --help` names `mempalace-code
+  wing-migration` in its usage line. The hash-pinned runner is unchanged, so open
+  1.14.2 receipts and qualification reports stay valid.
+- Bootstrap installs also link `~/.local/bin/mempalace-code-mcp`, and the
+  documented MCP launcher lookup resolves symlinks.
+- `update apply` keeps the `custom-models` extra, and missing-extra messages print
+  one install command for the running installation.
+- `fetch-model` reports the real cache size; a missing embedding cache makes
+  `mine` print one error naming `mempalace-code fetch-model` instead of a
+  traceback; `mempalace-code-mcp --help` shows its own name; an unreadable
+  `~/.mempalace/config.json` now produces a one-time warning instead of being
+  silently ignored.
+- The Dependency Audit issue lists each advisory once, and a clean run on `main`
+  closes it. GitHub Release notes now carry the version's changelog section, and
+  recovery releases compare against the previous published release.
+- The wheel and sdist include the upstream MemPalace MIT copyright and permission
+  notice in `NOTICE`.
+
+### Documentation
+
+- Corrected install, verification, backup/restore, offline, update, hook, search,
+  usage-rule, benchmark, and wing-migration documentation against the current CLI
+  and code: `init` does not create a palace; uv has no `tool inject`;
+  `config.json` examples are valid JSON; search is an exact flat scan with
+  cosine-equivalent ranking and no ANN index; `unknown_wing_room` is the real
+  filter error; README and PyPI links are absolute.
+- Documented the release runbook in execution order, Dependabot and pull-request
+  intake into the development trunk (never merged on public `main`), the CI gate
+  set for contributors (`.[dev,spellcheck,treesitter]`), and the legacy status of
+  the ChromaDB-based benchmark harnesses.
+- Corrected historical changelog entries (v1.0.0, v1.1.x, v1.2.0, v1.10.1,
+  v1.13.2) and upstream issue links.
+- Refreshed the reviewed upstream snapshot through `8c4865f7` (2026-09-26) and
+  aligned the README comparison, WHY_THIS_FORK, UPSTREAM_HARDENING, and
+  COMPARISON_GRAPHIFY with it.
+
+### Maintenance
+
+- Pinned GitHub Actions to actions/checkout v7.0.1, actions/cache v6.1.0,
+  actions/setup-python v7.0.0, and actions/setup-go v7.0.0 by full commit SHA.
+- Removed the inert project `.claude/settings.json` hooks, which read an
+  environment variable Claude Code never sets.
+- `release_public_read.py --version-tags` lists tags in release order; the
+  code-intelligence packet generator uses the canonical model cache and fails on
+  an MCP exhibit error; `embed_ab_bench.py` uses the maintained query set and
+  fails its LongMemEval gate loudly; `scripts/nuke_wing.py` requires `--yes` and
+  uses the configured palace.
+
 ## v1.14.2 — 2026-09-21
 
 Recovery release for the accumulated v1.14.1 changes. The immutable v1.14.1
@@ -208,10 +389,12 @@ build and publication.
 - Synchronized package, lockfile, README badge, Agent Plugins manifest, and
   generated quality scorecards on version 1.13.3.
 
-## v1.13.2 — 2026-08-11
+## v1.13.2 — 2026-08-11 (tagged, not published)
 
-Patch release for portable Agent Plugins packaging and collision-safe release
-identity checks.
+The tag remains immutable public evidence. The live upstream-drift check stopped
+publication before build, so no v1.13.2 package or GitHub Release exists. These
+changes (portable Agent Plugins packaging and collision-safe release identity
+checks) first shipped in v1.13.3.
 
 ### Added
 
@@ -487,7 +670,7 @@ Patch release for the local-first embedding path.
   search, and palace health.
 - Focused storage and CLI command suites plus the network-marked offline gate.
 
-## v1.10.1 — 2026-05-24
+## v1.10.1 — 2026-05-26
 
 Patch release after the v1.10.0 publish. Focus: real CLI/MCP smoke fixes,
 read-only no-embedder paths, and release-check tooling.
@@ -953,6 +1136,11 @@ Add `mempalace_explain_subsystem` MCP tool: combines semantic search with KG tra
 
 Add 4 architecture-oriented MCP tools for .NET type analysis: `mempalace_find_implementations`, `mempalace_find_references`, `mempalace_show_project_graph`, and `mempalace_show_type_dependencies`.
 
+## v1.2.0 — 2026-04-17
+
+Published release for the dated entries below, from SKILLS-HOOKS through
+FIX-LANCE-CORRUPT.
+
 ## 2026-04-17 · SKILLS-HOOKS
 
 Add Claude Code skills and hooks: 12 skills (`/start`, `/status`, `/verify`, `/palace-health`, `/task-plan`, `/task-hardening`, `/doc-refresh`, `/ship`, `/release`, `/entropy-gc`, `/mine`, `/bench`), 3 shared modules (mode-classification, task-state, commit-checkpoint), Codex review integration, pre-commit verification gate, and edit logging hooks.
@@ -997,10 +1185,19 @@ Auto-backup palace before risky operations: `safe_optimize` triggers a backup by
 
 Detect and recover from missing LanceDB fragment files: `safe_open_table` probes the table with a count query on open and rolls back to the last clean version automatically when fragment corruption is detected.
 
+## v1.1.1 — 2026-04-14
+
+Published patch release for the MINE-DEVOPS-INFRA entry below.
+
 ## 2026-04-14 · MINE-DEVOPS-INFRA
 
 Add DevOps/infrastructure file support to the miner: Terraform (`.tf`, `.tfvars`, `.hcl`), Dockerfiles, Makefiles, Helm templates (`.tpl`), Ansible Jinja2 templates (`.j2`, `.jinja2`), and general config files (`.conf`, `.cfg`, `.ini`) are now scanned and indexed.
 
+## v1.1.0 — 2026-04-14
+
+Published release for the dated entries below, from
+STORE-CHROMA-DELETE-WING-LIMIT through CODE-TREESITTER-INFRA.
+
 ## 2026-04-14 · STORE-CHROMA-DELETE-WING-LIMIT
 
 `ChromaStore.delete_wing` now calls `self.get()` instead of `self._col.get()`, so the `limit=10000` wrapper applies. Wings with more drawers than ChromaDB's default page size were silently partially deleted. (ChromaDB is deprecated; cleanup only.)
@@ -1061,10 +1258,11 @@ Tree-sitter AST-aware Python chunking: extracts function/class/method boundaries
 
 Tree-sitter optional infra: `.[treesitter]` extra, grammar download/cache, parser init, and automatic regex fallback when py-tree-sitter is absent or grammar unavailable.
 
-## v1.0.0 — 2026-04-12
+## v1.0.0 — 2026-04-13
 
 First public release of **mempalace-code**, a code-first fork of
-[milla-jovovich/mempalace](https://github.com/milla-jovovich/mempalace).
+[MemPalace/mempalace](https://github.com/MemPalace/mempalace) (then
+`milla-jovovich/mempalace`).
 
 ### Storage — LanceDB rewrite
 
@@ -1125,9 +1323,9 @@ First public release of **mempalace-code**, a code-first fork of
 
 ### Upstream issues addressed
 
-- [#469](https://github.com/milla-jovovich/mempalace/issues/469) — ChromaDB version-cliff data deletion → LanceDB, no version-cliff risk.
-- [#524](https://github.com/milla-jovovich/mempalace/issues/524) — Silent ONNX model download → explicit `mempalace init` + `fetch-model`.
-- [#27](https://github.com/milla-jovovich/mempalace/issues/27) — Unverifiable 100% R@5 claim → removed. AAAK "lossless" claim → labeled lossy.
+- [#469](https://github.com/MemPalace/mempalace/issues/469) — ChromaDB version-cliff data deletion → LanceDB, no version-cliff risk.
+- [#524](https://github.com/MemPalace/mempalace/issues/524) — Silent ONNX model download → explicit `mempalace init` + `fetch-model`.
+- [#27](https://github.com/MemPalace/mempalace/issues/27) — Unverifiable 100% R@5 claim → removed. AAAK "lossless" claim → labeled lossy.
 
 ### License