pubchem-finder 2.0.0__tar.gz

pubchem_finder-2.0.0/.github/CONTRIBUTING.md

@@ -0,0 +1,58 @@
+# Contributing
+
+Thanks for helping with PubChem Finder. The project is small; this page is the whole process.
+
+## Setup
+
+Install [uv](https://docs.astral.sh/uv/), then:
+
+```bash
+git clone https://github.com/MysterionRise/pubchem-finder.git
+cd pubchem-finder
+uv sync          # creates .venv with the project and dev tools (RDKit comes from PyPI wheels)
+make check       # ruff lint + format check, mypy --strict, pytest with coverage
+```
+
+Python 3.11, 3.12 and 3.13 are supported and tested in CI.
+
+## Workflow
+
+1. Branch from `master`.
+2. Make the change with tests. Unit tests go in `tests/unit/`, end-to-end flows in
+   `tests/integration/`, and tests that start the real MCP server process in `tests/e2e/`.
+   Tests must run offline; mark anything slow with `@pytest.mark.slow` and anything that
+   calls PubChem with `@pytest.mark.live`.
+3. Run `make check`. CI runs the same commands, so a green local run means a green PR.
+4. Open a pull request. Keep it focused and describe what changed and why.
+
+## What CI checks
+
+- **CI** (every push and pull request): lint, format and `mypy --strict` on Linux; the test
+  suite on Linux (Python 3.11, 3.12, 3.13), macOS and Windows; and a `package` job on all
+  three that builds the wheel, installs it into a clean environment with the newest
+  dependencies from PyPI, and runs `scripts/smoke.py` against the installed command.
+- **Live PubChem** (weekly and on demand): the `live` tests against the real PubChem
+  services. Run them locally with `PUBCHEM_LIVE=1 uv run pytest -m live`.
+- **Real-data end-to-end** (on demand, from the Actions tab): downloads real PubChem SDF
+  chunks, builds an index with substructure search, runs every CLI command and every MCP
+  tool against it and against PubChem, and writes build time, file sizes, memory and query
+  timings to the job summary. Locally: `uv run pubchem-finder index build --limit 1
+  --with-substructure`, then `uv run python scripts/smoke.py --home <data dir> --live`.
+
+## Code style
+
+- `ruff format` decides formatting; `ruff check` decides lint. Both run in CI and as
+  pre-commit hooks (`uv tool install pre-commit && uv run pre-commit install`).
+- `mypy --strict` must pass. RDKit is untyped; keep its objects behind small, typed
+  wrappers rather than sprinkling `Any` through the codebase.
+- Public functions get a short docstring. Comment the why, not the what.
+
+## Dependencies
+
+Add runtime dependencies with `uv add <package>` and dev tools with `uv add --dev <package>`,
+and commit the updated `uv.lock`. Do not add a dependency the code does not import yet.
+
+## Reporting problems
+
+Open an issue with the SMILES or input that triggers it, the command you ran, and the
+versions printed by `uv run python -c "import rdkit, pubchem_finder; print(rdkit.__version__, pubchem_finder.__version__)"`.

pubchem_finder-2.0.0/.github/ISSUE_TEMPLATE/bug_report.yml

@@ -0,0 +1,55 @@
+name: Bug report
+description: Something does not work as documented
+labels: [bug]
+body:
+  - type: textarea
+    id: what
+    attributes:
+      label: What happened
+      description: What you ran, what you expected, and what happened instead.
+    validations:
+      required: true
+  - type: textarea
+    id: command
+    attributes:
+      label: Command or MCP request
+      description: The exact command (or the tool call and arguments), and the SMILES or input that triggers it.
+      render: shell
+    validations:
+      required: true
+  - type: textarea
+    id: output
+    attributes:
+      label: Output
+      description: The full error or output. For the MCP server, include what your client shows and any stderr log.
+      render: text
+  - type: input
+    id: version
+    attributes:
+      label: pubchem-finder version
+      description: Output of `pubchem-finder --version`
+    validations:
+      required: true
+  - type: dropdown
+    id: install
+    attributes:
+      label: How you installed it
+      options:
+        - uvx / uv tool install (PyPI)
+        - uvx --from git+https://...
+        - pip
+        - from a clone (uv sync)
+    validations:
+      required: true
+  - type: input
+    id: platform
+    attributes:
+      label: Operating system and Python
+      placeholder: macOS 15 (Apple Silicon), Python 3.12
+    validations:
+      required: true
+  - type: input
+    id: client
+    attributes:
+      label: MCP client, if relevant
+      placeholder: Claude Desktop 1.x, Claude Code 2.x, ...

pubchem_finder-2.0.0/.github/ISSUE_TEMPLATE/config.yml

@@ -0,0 +1,5 @@
+blank_issues_enabled: true
+contact_links:
+  - name: Security issue
+    url: https://github.com/MysterionRise/pubchem-finder/security/advisories/new
+    about: Report security problems privately, not as a public issue.

pubchem_finder-2.0.0/.github/ISSUE_TEMPLATE/feature_request.yml

@@ -0,0 +1,16 @@
+name: Feature request
+description: Suggest a tool, command or improvement
+labels: [enhancement]
+body:
+  - type: textarea
+    id: problem
+    attributes:
+      label: What are you trying to do?
+      description: The chemistry question or workflow, and where the current tools fall short.
+    validations:
+      required: true
+  - type: textarea
+    id: idea
+    attributes:
+      label: What would help?
+      description: The tool, command or option you have in mind, with an example input and the result you would expect.

pubchem_finder-2.0.0/.github/PULL_REQUEST_TEMPLATE.md

@@ -0,0 +1,7 @@
+## What and why
+
+<!-- What this changes and the problem it solves. Link the issue if there is one. -->
+
+## How it was tested
+
+<!-- `make check` passes; anything else you ran (live tests, smoke script, a real index). -->

pubchem_finder-2.0.0/.github/dependabot.yml

@@ -0,0 +1,16 @@
+version: 2
+updates:
+  - package-ecosystem: uv
+    directory: /
+    schedule:
+      interval: weekly
+    groups:
+      python-dependencies:
+        patterns: ["*"]
+  - package-ecosystem: github-actions
+    directory: /
+    schedule:
+      interval: weekly
+    groups:
+      github-actions:
+        patterns: ["*"]

pubchem_finder-2.0.0/.github/workflows/ci.yml

@@ -0,0 +1,94 @@
+name: CI
+
+on:
+  push:
+    branches: [master, main, "claude/**", "feature/**"]
+  pull_request:
+    branches: [master, main]
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  test:
+    name: test (${{ matrix.os }}, ${{ matrix.python-version }})
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest]
+        python-version: ["3.11", "3.12", "3.13"]
+        include:
+          - os: macos-latest
+            python-version: "3.12"
+          - os: windows-latest
+            python-version: "3.12"
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install uv
+        uses: astral-sh/setup-uv@v6
+        with:
+          enable-cache: true
+          python-version: ${{ matrix.python-version }}
+
+      - name: Install project
+        run: uv sync --locked
+
+      - name: Lint
+        if: runner.os == 'Linux'
+        run: uv run ruff check .
+
+      - name: Format check
+        if: runner.os == 'Linux'
+        run: uv run ruff format --check .
+
+      - name: Type check
+        if: runner.os == 'Linux'
+        run: uv run mypy
+
+      - name: Test
+        run: uv run pytest -m "not slow" --cov --cov-report=term-missing --cov-report=xml
+
+      - name: Upload coverage report
+        if: runner.os == 'Linux' && matrix.python-version == '3.12'
+        uses: actions/upload-artifact@v4
+        with:
+          name: coverage-xml
+          path: coverage.xml
+
+  # Build the wheel, install it the way a user would (fresh environment, newest
+  # compatible dependencies from PyPI, not the lockfile) and smoke-test the
+  # installed command and MCP server.
+  package:
+    name: package (${{ matrix.os }})
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest, macos-latest, windows-latest]
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install uv
+        uses: astral-sh/setup-uv@v6
+        with:
+          python-version: "3.12"
+
+      - name: Build sdist and wheel
+        run: uv build
+
+      - name: Install the wheel into a clean environment
+        shell: bash
+        run: |
+          uv venv .smoke --python 3.12
+          uv pip install --python .smoke dist/*.whl
+
+      - name: Smoke test
+        shell: bash
+        run: |
+          if [ "$RUNNER_OS" = "Windows" ]; then bin=.smoke/Scripts; else bin=.smoke/bin; fi
+          "$bin/python" scripts/smoke.py --command "$bin/pubchem-finder"

pubchem_finder-2.0.0/.github/workflows/e2e-real.yml

@@ -0,0 +1,133 @@
+name: Real-data end-to-end
+
+# Builds a real index from PubChem's own SDF files and exercises every CLI
+# command and MCP tool against it and against the live PubChem services.
+# The job summary records build time, file sizes, memory and query timings.
+
+on:
+  workflow_dispatch:
+    inputs:
+      chunks:
+        description: "PubChem SDF chunks to index (about 500,000 compounds each)"
+        default: "1"
+
+permissions:
+  contents: read
+
+jobs:
+  e2e:
+    runs-on: ubuntu-latest
+    timeout-minutes: 90
+    env:
+      CHUNKS: ${{ inputs.chunks }}
+      PUBCHEM_FINDER_CACHE: "0"
+      ASPIRIN: "CC(=O)Oc1ccccc1C(=O)O"
+      INDOLE: "c1ccc2[nH]ccc2c1"
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install uv
+        uses: astral-sh/setup-uv@v6
+        with:
+          enable-cache: true
+          python-version: "3.12"
+
+      - name: Install project
+        run: uv sync --locked
+
+      - name: Prepare
+        run: |
+          echo "PUBCHEM_FINDER_HOME=$RUNNER_TEMP/pubchem-finder" >> "$GITHUB_ENV"
+          mkdir -p out
+
+      - name: Build the index from PubChem (download, fingerprints, substructure)
+        run: |
+          start=$(date +%s)
+          uv run pubchem-finder index build --limit "$CHUNKS" --with-substructure --workers 2 2>&1 | tee out/build.log
+          echo "BUILD_SECONDS=$(( $(date +%s) - start ))" >> "$GITHUB_ENV"
+          uv run pubchem-finder index status --json | tee out/status.json
+          jq -e '.state == "ready" and .substructure' out/status.json
+
+      - name: Local similarity and substructure search
+        run: |
+          uv run pubchem-finder search "$ASPIRIN" --threshold 0.7 --top 20 --json > out/search-aspirin.json
+          jq -e '.hits[0].cid == 2244 and .hits[0].similarity == 1' out/search-aspirin.json
+          uv run pubchem-finder search "$ASPIRIN" --threshold 0.6 --group-by-scaffold | tee out/search-scaffolds.txt
+          uv run pubchem-finder substructure "$INDOLE" --max 24 --json > out/substructure-indole.json
+          jq -e '.hits | length > 0' out/substructure-indole.json
+          uv run pubchem-finder substructure "[NX3][SX4](=O)(=O)c" --smarts --max 24 --json > out/substructure-sulfonamide.json
+          jq -e '.hits | length > 0' out/substructure-sulfonamide.json
+          printf '%s aspirin\nCN1C=NC2=C1C(=O)N(C(=O)N2C)C caffeine\nnot-a-smiles broken\n' "$ASPIRIN" > out/queries.smi
+          uv run pubchem-finder search -i out/queries.smi -o out/batch-hits.csv --threshold 0.7 --top 5
+          grep -q "^aspirin," out/batch-hits.csv
+
+      - name: Live PubChem commands
+        run: |
+          uv run pubchem-finder lookup caffeine --json > out/lookup-caffeine.json
+          jq -e '.compounds[0].cid == 2519' out/lookup-caffeine.json
+          uv run pubchem-finder pubchem synonyms 2244 --json > out/synonyms-2244.json
+          jq -e '.cas_numbers | index("50-78-2")' out/synonyms-2244.json
+          uv run pubchem-finder pubchem annotations 2244 --heading "GHS Classification" --json > out/ghs-2244.json
+          jq -e '.annotations | length > 0' out/ghs-2244.json
+          uv run pubchem-finder pubchem bioactivity 2244 --json > out/bioactivity-2244.json
+          jq -e '.n_assays > 0' out/bioactivity-2244.json
+          uv run pubchem-finder pubchem substructure "$INDOLE" --max 10 --json > out/pubchem-substructure-indole.json
+
+      - name: Pictures and triage
+        run: |
+          uv run pubchem-finder triage "$ASPIRIN" --json > out/triage-aspirin.json
+          uv run pubchem-finder draw $(jq -r '.hits[:8][].smiles' out/search-aspirin.json) \
+            --highlight diff --reference "$ASPIRIN" -o out/aspirin-analogues.png
+          uv run pubchem-finder draw $(jq -r '.hits[:8][].smiles' out/substructure-sulfonamide.json) \
+            --highlight smarts --smarts "[NX3][SX4](=O)(=O)c" -o out/sulfonamides.png
+
+      - name: Every MCP tool over stdio
+        run: uv run python scripts/smoke.py --home "$PUBCHEM_FINDER_HOME" --live
+
+      - name: Measure query timings
+        run: |
+          uv run python - <<'PY' | tee -a "$GITHUB_STEP_SUMMARY"
+          import json, os, time
+          from pathlib import Path
+          from pubchem_finder import service
+          from pubchem_finder.index.engine import LocalIndex
+
+          home = Path(os.environ["PUBCHEM_FINDER_HOME"])
+          status = json.loads(Path("out/status.json").read_text())
+          du = lambda p: sum(f.stat().st_size for f in p.rglob("*") if f.is_file())
+          mb = lambda n: f"{n / 1e6:,.0f} MB"
+          print("### Real-data index\n")
+          print("| | |\n|---|---|")
+          print(f"| PubChem chunks | {os.environ['CHUNKS']} |")
+          print(f"| Compounds indexed | {status['n_compounds']:,} ({status['n_failed']:,} skipped) |")
+          print(f"| Build (download, fingerprints, substructure, merge) | {int(os.environ['BUILD_SECONDS'])} s on 2 workers |")
+          print(f"| Similarity index file | {mb(status['size_bytes'])} |")
+          print(f"| Substructure index file | {mb(status['substructure_size_bytes'])} |")
+          print(f"| Chunk files (Parquet metadata + per-chunk fingerprints) | {mb(du(home / 'chunks'))} |")
+          print(f"| RAM for in-memory similarity / substructure | {mb(status['memory_bytes'])} / {mb(status['substructure_memory_bytes'])} |")
+          queries = {
+              "similarity, aspirin, top 20 >= 0.7": lambda idx: idx.search(
+                  service.parse_smiles(os.environ["ASPIRIN"]), 0.7, 20),
+              "substructure, indole, 50 hits": lambda idx: idx.substructure(
+                  service.parse_query(os.environ["INDOLE"]), 50, 100_000),
+              "substructure, aryl sulfonamide (SMARTS), 50 hits": lambda idx: idx.substructure(
+                  service.parse_query("[NX3][SX4](=O)(=O)c", "smarts"), 50, 100_000),
+          }
+          print("\n| Query | Mode | First call (loads index) | Then, per query |\n|---|---|---|---|")
+          for mode in ("memory", "disk"):
+              idx = LocalIndex(home, mode=mode)
+              for name, run in queries.items():
+                  t = time.perf_counter(); run(idx); first = time.perf_counter() - t
+                  t = time.perf_counter()
+                  for _ in range(5):
+                      run(idx)
+                  per = (time.perf_counter() - t) / 5
+                  print(f"| {name} | {mode} | {first * 1000:,.0f} ms | {per * 1000:,.1f} ms |")
+          PY
+
+      - name: Upload outputs
+        if: always()
+        uses: actions/upload-artifact@v4
+        with:
+          name: e2e-real-outputs
+          path: out/

pubchem_finder-2.0.0/.github/workflows/live.yml

@@ -0,0 +1,34 @@
+name: Live PubChem
+
+# The unit tests run offline against recorded responses. This workflow runs the
+# tests marked `live` against the real PubChem services, on demand and weekly,
+# so a change on PubChem's side shows up here before users hit it.
+
+on:
+  workflow_dispatch:
+  schedule:
+    - cron: "17 6 * * 1" # Mondays 06:17 UTC
+
+permissions:
+  contents: read
+
+jobs:
+  live:
+    runs-on: ubuntu-latest
+    timeout-minutes: 20
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install uv
+        uses: astral-sh/setup-uv@v6
+        with:
+          enable-cache: true
+          python-version: "3.12"
+
+      - name: Install project
+        run: uv sync --locked
+
+      - name: Live tests
+        env:
+          PUBCHEM_LIVE: "1"
+        run: uv run pytest -m live -v

pubchem_finder-2.0.0/.github/workflows/release.yml

@@ -0,0 +1,126 @@
+name: Release
+
+# Run from the Actions tab: "Release" -> "Run workflow".
+# With dry_run (the default) it only builds and tests the package. Without it, it
+# publishes to PyPI with trusted publishing, then tags the commit and creates the
+# GitHub Release with the CHANGELOG section as notes.
+#
+# One-time setup on pypi.org (Account -> Publishing -> Add a pending publisher):
+# project pubchem-finder, owner MysterionRise, repository pubchem-finder,
+# workflow release.yml, environment pypi.
+
+on:
+  workflow_dispatch:
+    inputs:
+      version:
+        description: "Version to release, without the v (must match pyproject.toml)"
+        required: true
+      dry_run:
+        description: "Build and test only; do not publish"
+        type: boolean
+        default: true
+
+permissions:
+  contents: read
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    env:
+      VERSION: ${{ inputs.version }}
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install uv
+        uses: astral-sh/setup-uv@v6
+        with:
+          python-version: "3.12"
+
+      - name: Check the version and changelog
+        run: |
+          project=$(uv run --no-project python -c "import tomllib; print(tomllib.load(open('pyproject.toml', 'rb'))['project']['version'])")
+          if [ "$project" != "$VERSION" ]; then
+            echo "::error::pyproject.toml says $project, not $VERSION"; exit 1
+          fi
+          if git ls-remote --exit-code --tags origin "refs/tags/v$VERSION" > /dev/null; then
+            echo "::error::tag v$VERSION already exists"; exit 1
+          fi
+          awk -v v="$VERSION" '
+            index($0, "## [" v "]") == 1 { on = 1; next }
+            on && /^## / { exit }
+            on { print }
+          ' CHANGELOG.md > release-notes.md
+          if [ ! -s release-notes.md ]; then
+            echo "::error::CHANGELOG.md has no section for $VERSION"; exit 1
+          fi
+
+      - name: Lint, type-check and test
+        run: |
+          uv sync --locked
+          uv run ruff check .
+          uv run ruff format --check .
+          uv run mypy
+          uv run pytest -m "not slow"
+
+      - name: Build sdist and wheel
+        run: uv build
+
+      - name: Smoke-test the wheel in a clean environment
+        run: |
+          uv venv .smoke --python 3.12
+          uv pip install --python .smoke dist/*.whl
+          .smoke/bin/python scripts/smoke.py --command .smoke/bin/pubchem-finder
+
+      - uses: actions/upload-artifact@v4
+        with:
+          name: dist
+          path: dist/
+
+      - uses: actions/upload-artifact@v4
+        with:
+          name: release-notes
+          path: release-notes.md
+
+  publish:
+    needs: build
+    if: ${{ !inputs.dry_run }}
+    runs-on: ubuntu-latest
+    environment:
+      name: pypi
+      url: https://pypi.org/project/pubchem-finder/${{ inputs.version }}/
+    permissions:
+      id-token: write # trusted publishing
+    steps:
+      - uses: actions/download-artifact@v4
+        with:
+          name: dist
+          path: dist/
+
+      - name: Publish to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1
+
+  github-release:
+    needs: publish
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    env:
+      VERSION: ${{ inputs.version }}
+      GH_TOKEN: ${{ github.token }}
+    steps:
+      - uses: actions/download-artifact@v4
+        with:
+          name: dist
+          path: dist/
+
+      - uses: actions/download-artifact@v4
+        with:
+          name: release-notes
+
+      - name: Tag and create the GitHub Release
+        run: |
+          gh release create "v$VERSION" dist/* \
+            --repo "$GITHUB_REPOSITORY" \
+            --target "$GITHUB_SHA" \
+            --title "v$VERSION" \
+            --notes-file release-notes.md

pubchem_finder-2.0.0/.gitignore

@@ -0,0 +1,30 @@
+# Python
+__pycache__/
+*.py[cod]
+*.egg-info/
+build/
+dist/
+
+# Environments
+.venv/
+venv/
+.env
+
+# Tooling caches
+.pytest_cache/
+.mypy_cache/
+.ruff_cache/
+.coverage
+.coverage.*
+htmlcov/
+coverage.xml
+
+# IDE
+.idea/
+.vscode/
+
+# Data and indexes (never commit PubChem dumps or fingerprint files)
+*.sdf.gz
+*.h5
+*.parquet
+/data/

pubchem_finder-2.0.0/.pre-commit-config.yaml

@@ -0,0 +1,14 @@
+repos:
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: v0.14.1
+    hooks:
+      - id: ruff
+        args: [--fix]
+      - id: ruff-format
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v5.0.0
+    hooks:
+      - id: end-of-file-fixer
+      - id: trailing-whitespace
+      - id: check-yaml
+      - id: check-toml

pubchem_finder-2.0.0/.python-version

@@ -0,0 +1 @@
+3.12

pubchem_finder-2.0.0/CHANGELOG.md

@@ -0,0 +1,55 @@
+# Changelog
+
+All notable changes to this project are documented here. The format follows
+[Keep a Changelog](https://keepachangelog.com/en/1.1.0/), and the project uses
+[Semantic Versioning](https://semver.org/).
+
+## [2.0.0] - 2026-10-22
+
+A complete rewrite. PubChem Finder is now a command-line tool, a Python library and an
+MCP server that lets Claude (or any MCP client) search PubChem by chemical structure,
+with a local index you can build on a laptop.
+
+### Added
+
+- **MCP server over stdio** (`pubchem-finder mcp`) with 17 tools for Claude Desktop,
+  Claude Code and other MCP clients.
+- **Local similarity search.** `index build` downloads PubChem's SDF files (or indexes
+  your own), fingerprints them with RDKit (ECFP4 by default) into an
+  [FPSim2](https://github.com/chembl/FPSim2) index with Parquet metadata, and searches
+  it exactly, in memory or from disk. Builds run in parallel and resume after an
+  interruption.
+- **Local substructure search** (`index build --with-substructure`, `substructure`):
+  an RDKit pattern-fingerprint screen, then an exact atom-by-atom match. SMILES or
+  SMARTS queries, results in CID order.
+- **PubChem knowledge tools:** lookup by name, CID, SMILES or InChIKey; PubChem's own
+  similarity, substructure and identity searches (hits re-scored with ECFP4); synonyms
+  and CAS numbers; cross-references (registry IDs, PubMed, patents); descriptions;
+  curated page sections such as GHS classification, pharmacology and toxicity, each
+  attributed to its source; bioassay summaries. Responses are cached for 7 days.
+- **Chemistry tools:** RDKit descriptors and Lipinski check, standardization, ECFP4 and
+  MACCS comparison, med-chem triage (QED, synthetic accessibility, Lipinski/Veber,
+  PAINS/Brenk/NIH alerts, Murcko scaffold), and structure drawings that highlight a
+  SMARTS pattern, what changed relative to a reference, or a similarity map.
+- **Batch mode** for `props`, `standardize`, `triage`, `search` and `substructure`:
+  `.smi`, `.csv` or `.tsv` in, CSV out, with an error column for bad rows.
+- **Testing:** the suite runs on Linux, macOS and Windows; the built wheel is installed
+  and smoke-tested on all three; live PubChem tests run weekly; a real-data workflow
+  builds an index from PubChem's own files and calls every tool.
+
+### Changed
+
+- Python 3.11 to 3.13, packaged with uv and hatchling, linted with ruff, type-checked
+  with mypy in strict mode.
+
+### Removed
+
+- The 2020 Indigo + Elasticsearch indexer, the cookiecutter template and the Docker
+  files. They remain in the git history before commit `06619d9`.
+
+## 1.x (2020)
+
+An Indigo + Elasticsearch indexer for PubChem SDF dumps. It was never released as a
+package.
+
+[2.0.0]: https://github.com/MysterionRise/pubchem-finder/releases/tag/v2.0.0

pubchem_finder-2.0.0/CITATION.cff

@@ -0,0 +1,18 @@
+cff-version: 1.2.0
+message: "If you use PubChem Finder in your work, please cite it as below, and cite PubChem for the data (see the README)."
+title: "PubChem Finder: search PubChem by chemical structure from the command line and from MCP clients"
+type: software
+authors:
+  - family-names: Perikov
+    given-names: Konstantin
+version: 2.0.0
+date-released: "2026-10-22"
+license: MIT
+repository-code: "https://github.com/MysterionRise/pubchem-finder"
+keywords:
+  - cheminformatics
+  - PubChem
+  - RDKit
+  - similarity search
+  - substructure search
+  - Model Context Protocol