Add codespell support with configuration and typo fixes

.github/workflows/codespell.yml

@@ -0,0 +1,23 @@
+# Codespell configuration is within pyproject.toml
+---
+name: Codespell
+
+on:
+  push:
+    branches: [v8]
+  pull_request:
+    branches: [v8]
+
+permissions:
+  contents: read
+
+jobs:
+  codespell:
+    name: Check for spelling errors
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v6
+      - name: Codespell
+        uses: codespell-project/actions-codespell@8f01853be192eb0f849a5c7d721450e7a467c579  # v2.2

.pre-commit-config.yaml

@@ -20,3 +20,11 @@ repos:
     hooks:
       - id: ruff
         args: ["--config", "pyproject.toml"]
+
+  - repo: https://github.com/codespell-project/codespell
+    # Configuration for codespell is in pyproject.toml
+    rev: v2.4.2
+    hooks:
+    - id: codespell
+      additional_dependencies:
+      - tomli; python_version<'3.11'

docs/superpowers/plans/2026-05-04-incremental-updates-dedup.md

@@ -101,7 +101,7 @@ def test_entropy_empty_string():
 
 def test_shingles_produces_trigrams():
     s = _shingles("hello")
-    assert "hel" in s
+    assert "hel" in s  # codespell:ignore hel
     assert "ell" in s
     assert "llo" in s
 

graphify/extract.py

@@ -5510,7 +5510,7 @@ def _has_import_evidence(candidate_id: str) -> bool:
             # `indirect_call` and ONLY when the target is a real callable def —
             # never a same-named data symbol. Stays INFERRED even with import
             # evidence: the name is referenced as a value here, not invoked. Dedup
-            # is call-aware (an existing direct `calls` edge pre-empts it; a benign
+            # is call-aware (an existing direct `calls` edge preempts it; a benign
             # `imports` edge to the same symbol does NOT suppress it).
             if tgt != caller and (caller, tgt) not in call_like_pairs and tgt in callable_nids and tgt not in class_nids:
                 call_like_pairs.add((caller, tgt))

graphify/extractors/dart.py

@@ -480,7 +480,7 @@ def _find_matching_brace(text: str, start_pos: int) -> int:
                     add_node(bloc_nid, bloc_name, source_file=None)
                     add_edge(nid, bloc_nid, "references", context="bloc_lookup")
 
-            # Universal Navigation Patters (GoRouter, AutoRoute, Navigator)
+            # Universal Navigation Patterns (GoRouter, AutoRoute, Navigator)
             for nm in re.finditer(r"\b(?:go|push|goNamed|pushNamed|replace|replaceNamed)\s*\(\s*(?:context\s*,\s*)?['\"]([a-zA-Z0-9_/?=&%-]+)['\"]", func_body):
                 route_path = nm.group(1)
                 route_nid = _make_id("route", route_path.replace("/", "_").replace("?", "_").replace("=", "_").replace("&", "_"))

graphify/llm.py

@@ -470,7 +470,7 @@ def _thinking_disabled_via_env() -> bool:
 Edge direction rule — source is always the ACTOR, target is the ACTED-UPON:
 - calls: source = the function/method that CONTAINS the call site; target = the function/method BEING CALLED. Never reverse this.
 - imports/references: source = the file/entity that imports or references; target = the thing imported or referenced.
-- implements/inherits: source = the subclass/implementor; target = the base class/interface.
+- implements/inherits: source = the subclass/implementer; target = the base class/interface.
 
 Hyperedges: if 3 or more nodes clearly participate together in a shared concept, flow, or pattern that is not captured by pairwise edges alone, add a hyperedge to the top-level `hyperedges` array (e.g. all classes implementing one protocol, all functions in one auth flow even if they don't all call each other, all concepts from a paper section forming one coherent idea). Use sparingly — only when the group relationship adds information beyond the pairwise edges. Maximum 3 hyperedges per chunk.
 

graphify/tree_html.py

@@ -1,6 +1,6 @@
 """tree_html — emit a D3 v7 collapsible-tree HTML view of a graph.
 
-A self-contained printable / browseable tree-of-modules view
+A self-contained printable / browsable tree-of-modules view
 intended to complement the existing force-directed ``graph.html``.
 Key visual elements:
 

pyproject.toml

@@ -150,3 +150,26 @@ select = ["E9", "F63", "F7", "F82"]
 include = ["graphify", "tests"]
 pythonVersion = "3.10"
 typeCheckingMode = "basic"
+
+[tool.codespell]
+# Ref: https://github.com/codespell-project/codespell#using-a-config-file
+# worked/ contains sample outputs (graph.json/html snapshots from real repos)
+# whose truncated string identifiers extracted from source are not typos in
+# the codebase itself. translations/ (basename glob, matches docs/translations/)
+# hosts foreign-language README variants that codespell (English-only) misreads.
+skip = '.git,.gitignore,.gitattributes,*.svg,*.lock,worked,translations'
+check-hidden = true
+# Per-word rationale for ignore-words-list (single TOML string, per-entry
+# comments are not possible inside the value):
+#   unparseable - consistent alternate spelling used across graphify/ (35x)
+#   bloc        - Dart BLoC pattern (Business Logic Component), not "block"
+#   inout       - SQL PROCEDURE param mode / Verilog keyword (not "input")
+#   dependant   - FastAPI class name shown in README example output
+#   inh         - test-local variable holding "inherits" edge set (tests/)
+#   datas       - matches edge_datas() function name from graphify.build
+#   nd          - short variable for graph node data (wiki.py, engine.py)
+#   bu          - short variable for Bedrock usage dict (llm.py)
+#   anc         - short variable for ancestor path (resolution.py)
+#   ans         - short variable for answer file path (tests/test_reflect.py)
+#   oder,ist    - German stopwords in serve.py's stopword list
+ignore-words-list = 'unparseable,bloc,inout,dependant,inh,datas,nd,bu,anc,ans,oder,ist'

tests/test_dedup.py

@@ -20,7 +20,7 @@ def test_entropy_empty_string():
 
 def test_shingles_produces_trigrams():
     s = _shingles("hello")
-    assert "hel" in s
+    assert "hel" in s  # codespell:ignore hel
     assert "ell" in s
     assert "llo" in s