v0.5.2: Fix fragile tuple default, stale glossary formula, CLI subcommand count

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: IronAdamant

CHANGELOG.md

@@ -5,6 +5,14 @@ All notable changes to Chisel are documented in this file.
 Format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 This project uses [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.5.2] - 2026-03-22
+
+### Fixed
+
+- `metrics.py`: Fragile tuple default `(None,)` in `compute_co_changes()` changed to `(None, None)` for consistency with the 2-element tuple structure
+- `wiki-local/glossary.md`: Stale 4-component risk formula updated to current 5-component formula (0.35/0.25/0.2/0.1/0.1 with test_instability)
+- `CLAUDE.md`: Corrected CLI subcommand count from 18 to 17
+
 ## [0.5.1] - 2026-03-22
 
 ### Fixed

CLAUDE.md

@@ -14,7 +14,7 @@ chisel/
   metrics.py        — Pure computation: churn scoring, ownership aggregation, co-change detection. _parse_iso_date shared utility.
   test_mapper.py    — Test file discovery, framework detection, dependency extraction, edge building.
   impact.py         — Impact analysis, risk scoring, stale test detection, reviewer suggestions. Caches failure rates.
-  cli.py            — argparse CLI (18 subcommands). _run_tool() shared handler. Entry point: chisel.cli:main
+  cli.py            — argparse CLI (17 subcommands). _run_tool() shared handler. Entry point: chisel.cli:main
   schemas.py        — JSON Schema definitions for all 15 tools + dispatch table. Shared by HTTP and stdio servers.
   mcp_server.py     — HTTP MCP server (GET /tools, /health, POST /call). ThreadedHTTPServer. dispatch_tool() shared by both servers.
   mcp_stdio.py      — stdio MCP server (requires optional 'mcp' package). _configure_server() for engine lifecycle mgmt.

COMPLETE_PROJECT_DOCUMENTATION.md

@@ -2,7 +2,7 @@
 
 Test impact analysis and code intelligence for LLM agents. Zero external dependencies.
 
-**Version:** 0.5.1
+**Version:** 0.5.2
 **PyPI:** `chisel-test-impact`
 **License:** MIT
 **Python:** >= 3.9

chisel/__init__.py

@@ -1 +1 @@
-__version__ = "0.5.1"
+__version__ = "0.5.2"

chisel/metrics.py

@@ -196,7 +196,7 @@ def compute_co_changes(commits, min_count=3):
                 "file_a": a,
                 "file_b": b,
                 "co_commit_count": count,
-                "last_co_commit": pair_last.get((a, b), (None,))[0],
+                "last_co_commit": pair_last.get((a, b), (None, None))[0],
             })
 
     result.sort(key=lambda x: x["co_commit_count"], reverse=True)

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "chisel-test-impact"
-version = "0.5.1"
+version = "0.5.2"
 description = "Test impact analysis and code intelligence for LLM agents"
 readme = "README.md"
 requires-python = ">=3.9"

wiki-local/glossary.md

@@ -47,7 +47,7 @@ The role assigned to entries returned by the `ownership` tool. Determined by `gi
 A test edge whose `code_id` points to a code unit that no longer exists in the `code_units` table. This is possible because SQLite foreign key enforcement is intentionally disabled. Orphaned references are the mechanism by which stale test detection works.
 
 **Risk score**
-A composite metric indicating how risky it is to change a file. Formula: `0.4 * churn + 0.3 * coupling_breadth + 0.2 * (1 - test_coverage) + 0.1 * author_concentration`. Each component is normalized to the 0-1 range. Higher values indicate higher risk. Computed by `ImpactAnalyzer.compute_risk_score()`.
+A composite metric indicating how risky it is to change a file. Formula: `0.35 * churn + 0.25 * coupling + 0.2 * coverage_gap + 0.1 * author_concentration + 0.1 * test_instability`. Each component is normalized to the 0-1 range. Higher values indicate higher risk. Computed by `ImpactAnalyzer.compute_risk_score()`.
 
 **RWLock (read-write lock)**
 A concurrency primitive in `rwlock.py` that allows multiple concurrent readers or one exclusive writer. Used by `ChiselEngine` to protect storage access: `tool_*()` read methods acquire a read lock, while `analyze()` and `update()` acquire a write lock.