chore(release): v1.1.1 (#12)

Patch release bundling every main-branch delta since the v1.1.0
marketplace artifact, headlined by the claude --print CWD pollution
fix from PR #11.

- Bump .claude-plugin/plugin.json to 1.1.1 so the plugin marketplace
  cache re-fetches the patched code on next refresh.
- Add scripts/check_summariser_pollution.py + tests as the regression
  detector for the cwd-pollution bug. Stdlib only, exits non-zero with
  a file list if any .jsonl outside ~/.claude/projects/-root--lossless-code--cli-cwd
  contains the summariser prompt.
- Expand CHANGELOG.md to cover the full v1.1.0 -> v1.1.1 surface
  (cwd fix, fingerprint feature, db package split, BM25 eviction,
  session filtering, circuit breaker, claude-cli provider, etc).
- Document the second deployment gap in the cwd-pollution learning
  doc: marketplace cache is a separate artifact from the manual
  install, so install.sh alone cannot deliver the fix to plugin
  users.

Author: GodsBoy

.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "lossless-code",
   "description": "DAG-based lossless context management for Claude Code. Every message from every session is persisted in a SQLite vault with FTS5 search. Summaries form a directed acyclic graph - nothing is ever lost.",
-  "version": "1.1.0",
+  "version": "1.1.1",
   "author": {
     "name": "GodsBoy",
     "email": "dhuysamen@gmail.com"

CHANGELOG.md

@@ -1,8 +1,29 @@
 # Changelog
 
-## Unreleased
+## v1.1.1 — 2026-04-14
 
-### Added — fingerprint file context (default off)
+Patch release that ships every main-branch delta accumulated since the
+v1.1.0 marketplace artifact, headlined by the `claude --print` CWD fix
+that closes the `claude --resume` pollution bug. Marketplace users
+should upgrade to pick up the fix; manual installs already covered by
+re-running `install.sh`.
+
+### Fixed
+
+- **`claude --print` subprocess CWD pollution** (PR #11). The
+  summariser's `claude --print` calls now pin `cwd=` to
+  `~/.lossless-code/.cli-cwd`, so Claude Code files internal
+  summarisation sessions in their own bucket instead of polluting the
+  user's `claude --resume` list with hundreds of "Summarise the
+  following conversation turns concisely…" entries. Includes
+  `scripts/check_summariser_pollution.py` regression detector.
+- **Installer + README sync for fingerprint feature** (PR #10).
+  `install.sh` now copies the new PreToolUse / PostToolUse hooks and
+  the `scripts/file_context.py`, `scripts/hook_store_tool_call.py`
+  files added by the fingerprint feature; README documents the new
+  `fileContextEnabled` flag and the hook surface.
+
+### Added — fingerprint file context (default off, PR #9)
 
 - **`fileContextEnabled` config flag** (default `false`) gates all new
   behavior. Flip to `true` in `~/.lossless-code/config.json` to opt in.
@@ -30,3 +51,42 @@
   from a file path to the recent summaries that reference it.
 - **`lcc status`** surfaces tagged message count, distinct tagged files,
   and cached fingerprint count when the flag is on.
+
+### Changed
+
+- **`scripts/db.py` split into `scripts/db/` package** (PR #8). Single
+  745-line module replaced with focused submodules: `config`,
+  `schema`, `messages`, `summaries`, `sessions`, `embeddings`,
+  `search`, `dream_log`. Public import surface is preserved via
+  `scripts/db/__init__.py`; existing call sites unchanged.
+- **BM25 prompt-aware context eviction** (PR #7). Context injection now
+  scores candidate summaries against the live prompt with BM25 instead
+  of a flat recency cut, so high-signal older summaries survive
+  eviction when the prompt actually references them.
+
+### Added — earlier in the v1.1.0 → v1.1.1 window
+
+- **Session filtering for lossless-claw v0.7.0 parity**: stateless
+  session gate, stop-hook filtering, plus DB-layer tests covering
+  pattern matching and filter behaviour.
+- **Circuit breaker + dynamic chunk sizing for summarisation**:
+  protects the summariser from runaway provider failures and adapts
+  chunk size to the active provider's context window.
+- **Vault bloat protection**: summary size caps and dream pagination
+  prevent unbounded growth on long-running vaults.
+- **`claude-cli` provider** for Claude Max / Pro subscription users
+  (no API key path). `ANTHROPIC_API_KEY` is stripped from the
+  subprocess env so the CLI falls back to OAuth.
+- **Custom Anthropic-compatible providers** (MiniMax and others) via
+  `openaiBaseUrl`, plus provider auto-detection and structured error
+  handling.
+- **`dream --status` flag** for inspecting dream-cycle state.
+- **`scripts/check_summariser_pollution.py`** stdlib regression detector
+  for the cwd pollution bug. Walks `~/.claude/projects/`, skips the
+  legitimate `-root--lossless-code--cli-cwd` bucket, exits non-zero
+  with a file list if any other bucket contains a polluting `.jsonl`.
+
+## v1.1.0 — semantic search hybrid
+
+- Hybrid semantic search combining FTS5 BM25 with vector similarity via
+  reciprocal rank fusion. Cached `fastembed.TextEmbedding` instance.

scripts/check_summariser_pollution.py

@@ -0,0 +1,68 @@
+#!/usr/bin/env python3
+"""Detect summariser CWD pollution in Claude Code project session buckets.
+
+Walks ``~/.claude/projects/`` (override with ``LOSSLESS_CHECK_PROJECTS_DIR``),
+skipping the legitimate ``-root--lossless-code--cli-cwd`` bucket. Any ``.jsonl``
+file containing the lossless-code summariser prompt outside that bucket is a
+regression of the cwd-pinning fix and causes a non-zero exit.
+
+Exit codes:
+    0  no pollution found (or projects dir missing)
+    1  one or more polluting files found (printed to stdout)
+"""
+
+from __future__ import annotations
+
+import os
+import sys
+from pathlib import Path
+
+LEGIT_BUCKET = "-root--lossless-code--cli-cwd"
+NEEDLE = (
+    b"Summarise the following conversation turns concisely, "
+    b"preserving all key decisions, facts, file paths, commands"
+)
+SCAN_BYTES = 8192
+
+
+def projects_dir() -> Path:
+    override = os.environ.get("LOSSLESS_CHECK_PROJECTS_DIR")
+    if override:
+        return Path(override)
+    return Path.home() / ".claude" / "projects"
+
+
+def file_is_polluting(path: Path) -> bool:
+    try:
+        with path.open("rb") as fh:
+            head = fh.read(SCAN_BYTES)
+    except OSError:
+        return False
+    return NEEDLE in head
+
+
+def find_polluting(root: Path) -> list[Path]:
+    if not root.is_dir():
+        return []
+    hits: list[Path] = []
+    for bucket in sorted(root.iterdir()):
+        if not bucket.is_dir() or bucket.name == LEGIT_BUCKET:
+            continue
+        for entry in bucket.rglob("*.jsonl"):
+            if entry.is_file() and file_is_polluting(entry):
+                hits.append(entry)
+    return hits
+
+
+def main() -> int:
+    hits = find_polluting(projects_dir())
+    if not hits:
+        return 0
+    print(f"Found {len(hits)} polluting jsonl file(s) outside {LEGIT_BUCKET}:")
+    for path in hits:
+        print(f"  {path}")
+    return 1
+
+
+if __name__ == "__main__":
+    sys.exit(main())

tests/test_check_summariser_pollution.py

@@ -0,0 +1,85 @@
+#!/usr/bin/env python3
+"""Tests for scripts/check_summariser_pollution.py regression detector."""
+
+import os
+import subprocess
+import sys
+import tempfile
+import unittest
+from pathlib import Path
+
+SCRIPT = Path(__file__).resolve().parent.parent / "scripts" / "check_summariser_pollution.py"
+
+POLLUTING_LINE = (
+    '{"type":"user","message":{"role":"user","content":"Summarise the following '
+    'conversation turns concisely, preserving all key decisions, facts, file paths, '
+    'commands and outputs"}}\n'
+)
+CLEAN_LINE = '{"type":"user","message":{"role":"user","content":"hello world"}}\n'
+
+
+def _run(projects_dir: Path) -> subprocess.CompletedProcess:
+    env = dict(os.environ)
+    env["LOSSLESS_CHECK_PROJECTS_DIR"] = str(projects_dir)
+    return subprocess.run(
+        [sys.executable, str(SCRIPT)],
+        env=env,
+        capture_output=True,
+        text=True,
+    )
+
+
+def _write_jsonl(path: Path, body: str) -> None:
+    path.parent.mkdir(parents=True, exist_ok=True)
+    path.write_text(body)
+
+
+class TestCheckSummariserPollution(unittest.TestCase):
+    def setUp(self):
+        self.tmp = tempfile.TemporaryDirectory()
+        self.addCleanup(self.tmp.cleanup)
+        self.projects = Path(self.tmp.name)
+
+    def test_zero_polluting_three_clean_exits_0(self):
+        bucket = self.projects / "-root-foo"
+        for i in range(3):
+            _write_jsonl(bucket / f"clean{i}.jsonl", CLEAN_LINE)
+        result = _run(self.projects)
+        self.assertEqual(result.returncode, 0, result.stdout + result.stderr)
+
+    def test_polluting_files_exit_1_and_listed(self):
+        bucket = self.projects / "-root-foo"
+        _write_jsonl(bucket / "bad1.jsonl", POLLUTING_LINE)
+        _write_jsonl(bucket / "bad2.jsonl", POLLUTING_LINE)
+        _write_jsonl(bucket / "ok.jsonl", CLEAN_LINE)
+        result = _run(self.projects)
+        self.assertEqual(result.returncode, 1)
+        self.assertIn("bad1.jsonl", result.stdout)
+        self.assertIn("bad2.jsonl", result.stdout)
+        self.assertNotIn("ok.jsonl", result.stdout)
+
+    def test_polluting_in_cli_cwd_bucket_ignored(self):
+        bucket = self.projects / "-root--lossless-code--cli-cwd"
+        _write_jsonl(bucket / "legit.jsonl", POLLUTING_LINE)
+        result = _run(self.projects)
+        self.assertEqual(result.returncode, 0, result.stdout + result.stderr)
+
+    def test_empty_jsonl_is_clean(self):
+        bucket = self.projects / "-root-foo"
+        _write_jsonl(bucket / "empty.jsonl", "")
+        result = _run(self.projects)
+        self.assertEqual(result.returncode, 0)
+
+    def test_malformed_jsonl_does_not_crash(self):
+        bucket = self.projects / "-root-foo"
+        _write_jsonl(bucket / "junk.jsonl", "not json at all\n{broken")
+        result = _run(self.projects)
+        self.assertEqual(result.returncode, 0)
+
+    def test_missing_projects_dir_exits_0(self):
+        result = _run(self.projects / "does-not-exist")
+        self.assertEqual(result.returncode, 0)
+
+
+if __name__ == "__main__":
+    unittest.main()