refactor

Name: refactor
Author: Borda

23stars3forks

How to install refactor?

borda/ai-rig/refactor

$npx -y skills add Borda/AI-Rig --skill refactor

Installs into the current project.

Summary

Test-first refactoring — audit coverage, add characterization tests, apply changes with safety net, run quality stack and review loop. TRIGGER when: user wants to restructure existing Python code without changing behaviour; phrases: \"refactor X\", \"clean up Y\", \"extract Z\",

Files · 1

View on GitHub

SKILL.md

1	<objective>
2
3	Test-first refactoring. Audit coverage, add characterization tests if missing, apply changes with safety net.
4
5	NOT for:
6	- bug fixes (use `/develop:fix`)
7	- new features (use `/develop:feature`)
8	- `.claude/` config changes (use `/foundry:manage` (requires foundry plugin))
9	- non-Python projects (JS/TS/Go/Rust) — toolchain assumes pytest; use language-native toolchain instead
10	- mixed refactor+feature tasks — run /develop:refactor first, then /develop:feature; do not attempt both in single skill run
11
12	Quality stack (Branch Safety Guard, Codex Pre-pass, Progressive Review) requires `foundry` plugin; when absent, Step 5 quality stack skipped with a visible warning — output lower quality but workflow still completes.
13
14	</objective>
15
16	<constants>
17
18	- MAX_INNER_CYCLES: 5 (change-test cycles per outer session — Step 4 safety break)
19
20	</constants>
21
22	<compaction>
23
24	Key boundary: end of Step 2 — coverage audit complete, before characterization test writing in Step 3.
25	Second boundary: end of Step 4 — refactor edits applied, before review stack in Step 5.
26	Preserve at boundary 1: dev-dir, target path, coverage audit summary, plan-file, --keep items.
27	Preserve at boundary 2: dev-dir, changed files list, test outcomes.
28
29	</compaction>
30
31	<workflow>
32
33	<!-- Agent Resolution: resolved at runtime via $_DEV_SHARED; source at develop/skills/_shared/agent-resolution.md (installed path) -->
34
35	## Agent Resolution
36
37	```bash
38	_PATHS=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_develop}/bin/dev_shared_resolve.py" --foundry 2>/dev/null) # timeout: 5000
39	_DEV_SHARED=$(echo "$_PATHS" \| head -1)
40	_FOUNDRY_SHARED=$(echo "$_PATHS" \| tail -1)
41	[ -z "$_FOUNDRY_SHARED" ] && _FOUNDRY_SHARED="plugins/cc_foundry/skills/_shared"
42	# loads: compaction-contract.md
43	cat "$_DEV_SHARED/agent-resolution.md"
44	```
45
46	Contains: foundry check + fallback table. If foundry not installed: substitute each `foundry:X` with `general-purpose` per table. Agents skill uses: `foundry:sw-engineer`, `foundry:qa-specialist`, `foundry:linting-expert`, `foundry:challenger`.
47
48	```bash
49	_DEV_SHARED=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_develop}/bin/dev_shared_resolve.py" 2>/dev/null) # timeout: 5000
50	cat "$_DEV_SHARED/task-hygiene.md"
51	```
52
53	<!-- Reference only — execution-dead at runtime; included for agent behavioral context -->
54	## Anti-Rationalizations
55
56	\| Temptation \| Reality \|
57	\| --- \| --- \|
58	\| "The code is simple enough — I can skip characterization tests" \| No safety net = no proof behavior unchanged. Characterization tests only proof. \|
59	\| "I'll fix this adjacent bug while I'm in here" \| Scope creep conflates history. Adjacent bugs go in Follow-up, not this session. \|
60	\| "The tests are too brittle — I'll refactor them as well" \| Refactoring tests + prod code simultaneously makes regressions unattributable. Fix tests first, separate pass. \|
61	\| "I know the codebase — no need for coverage audit" \| Untested edge cases = most common refactoring breakage. Audit finds what you don't know you don't know. \|
62	\| "This is a small change — Step 4's max-5 cycles are overkill" \| Simple changes = simple test loops. Guard costs nothing when unneeded; prevents runaway sessions when it is. \|
63
64	## Project Detection
65
66	```bash
67	_DEV_SHARED=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_develop}/bin/dev_shared_resolve.py" 2>/dev/null) # timeout: 5000
68	cat "$_DEV_SHARED/runner-detection.md"
69	```
70	Sets `$TEST_CMD` (full suite) and `$PYTEST_CMD` (pytest flags). Run at skill start.
71
72	Optional `--plan <path>`: if `$ARGUMENTS` contains `--plan <path>` (at any position), read plan file first. Extract `Affected files`, `Risks`, `Suggested approach` — use to inform Step 1 scope analysis. Skip redundant codebase exploration for already-classified files. Store plan path as `PLAN_FILE`.
73
74	```bash
75	_DEV_SHARED=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_develop}/bin/dev_shared_resolve.py" 2>/dev/null) # timeout: 5000
76	cat "$_DEV_SHARED/preflight-helpers.md"
77	```
78	Execute --plan path extraction; sets `$PLAN_FILE`.
79
80	Checkpoint init: run `DEV_DIR=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_develop}/bin/dev_run_dir.py" 2>/dev/null) #

People who liked this also viewed

6 picks

Type

refactor

23stars3forks

View on GitHub

How to install refactor?

borda/ai-rig/refactor

$npx -y skills add Borda/AI-Rig --skill refactor

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	<objective>
2
3	Test-first refactoring. Audit coverage, add characterization tests if missing, apply changes with safety net.
4
5	NOT for:
6	- bug fixes (use `/develop:fix`)
7	- new features (use `/develop:feature`)
8	- `.claude/` config changes (use `/foundry:manage` (requires foundry plugin))
9	- non-Python projects (JS/TS/Go/Rust) — toolchain assumes pytest; use language-native toolchain instead
10	- mixed refactor+feature tasks — run /develop:refactor first, then /develop:feature; do not attempt both in single skill run
11
12	Quality stack (Branch Safety Guard, Codex Pre-pass, Progressive Review) requires `foundry` plugin; when absent, Step 5 quality stack skipped with a visible warning — output lower quality but workflow still completes.
13
14	</objective>
15
16	<constants>
17
18	- MAX_INNER_CYCLES: 5 (change-test cycles per outer session — Step 4 safety break)
19
20	</constants>
21
22	<compaction>
23
24	Key boundary: end of Step 2 — coverage audit complete, before characterization test writing in Step 3.
25	Second boundary: end of Step 4 — refactor edits applied, before review stack in Step 5.
26	Preserve at boundary 1: dev-dir, target path, coverage audit summary, plan-file, --keep items.
27	Preserve at boundary 2: dev-dir, changed files list, test outcomes.
28
29	</compaction>
30
31	<workflow>
32
33	<!-- Agent Resolution: resolved at runtime via $_DEV_SHARED; source at develop/skills/_shared/agent-resolution.md (installed path) -->
34
35	## Agent Resolution
36
37	```bash
38	_PATHS=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_develop}/bin/dev_shared_resolve.py" --foundry 2>/dev/null) # timeout: 5000
39	_DEV_SHARED=$(echo "$_PATHS" \| head -1)
40	_FOUNDRY_SHARED=$(echo "$_PATHS" \| tail -1)
41	[ -z "$_FOUNDRY_SHARED" ] && _FOUNDRY_SHARED="plugins/cc_foundry/skills/_shared"
42	# loads: compaction-contract.md
43	cat "$_DEV_SHARED/agent-resolution.md"
44	```
45
46	Contains: foundry check + fallback table. If foundry not installed: substitute each `foundry:X` with `general-purpose` per table. Agents skill uses: `foundry:sw-engineer`, `foundry:qa-specialist`, `foundry:linting-expert`, `foundry:challenger`.
47
48	```bash
49	_DEV_SHARED=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_develop}/bin/dev_shared_resolve.py" 2>/dev/null) # timeout: 5000
50	cat "$_DEV_SHARED/task-hygiene.md"
51	```
52
53	<!-- Reference only — execution-dead at runtime; included for agent behavioral context -->
54	## Anti-Rationalizations
55
56	\| Temptation \| Reality \|
57	\| --- \| --- \|
58	\| "The code is simple enough — I can skip characterization tests" \| No safety net = no proof behavior unchanged. Characterization tests only proof. \|
59	\| "I'll fix this adjacent bug while I'm in here" \| Scope creep conflates history. Adjacent bugs go in Follow-up, not this session. \|
60	\| "The tests are too brittle — I'll refactor them as well" \| Refactoring tests + prod code simultaneously makes regressions unattributable. Fix tests first, separate pass. \|
61	\| "I know the codebase — no need for coverage audit" \| Untested edge cases = most common refactoring breakage. Audit finds what you don't know you don't know. \|
62	\| "This is a small change — Step 4's max-5 cycles are overkill" \| Simple changes = simple test loops. Guard costs nothing when unneeded; prevents runaway sessions when it is. \|
63
64	## Project Detection
65
66	```bash
67	_DEV_SHARED=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_develop}/bin/dev_shared_resolve.py" 2>/dev/null) # timeout: 5000
68	cat "$_DEV_SHARED/runner-detection.md"
69	```
70	Sets `$TEST_CMD` (full suite) and `$PYTEST_CMD` (pytest flags). Run at skill start.
71
72	Optional `--plan <path>`: if `$ARGUMENTS` contains `--plan <path>` (at any position), read plan file first. Extract `Affected files`, `Risks`, `Suggested approach` — use to inform Step 1 scope analysis. Skip redundant codebase exploration for already-classified files. Store plan path as `PLAN_FILE`.
73
74	```bash
75	_DEV_SHARED=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_develop}/bin/dev_shared_resolve.py" 2>/dev/null) # timeout: 5000
76	cat "$_DEV_SHARED/preflight-helpers.md"
77	```
78	Execute --plan path extraction; sets `$PLAN_FILE`.
79
80	Checkpoint init: run `DEV_DIR=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_develop}/bin/dev_run_dir.py" 2>/dev/null) #

People who liked this also viewed

6 picks

Type