@rune-kit/rune 2.10.0 → 2.11.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (205) hide show
  1. package/LICENSE +21 -21
  2. package/README.md +8 -6
  3. package/commands/rune.md +168 -168
  4. package/contexts/dev.md +34 -34
  5. package/contexts/research.md +43 -43
  6. package/contexts/review.md +55 -55
  7. package/extensions/ai-ml/PACK.md +88 -88
  8. package/extensions/ai-ml/skills/ai-agents.md +172 -172
  9. package/extensions/ai-ml/skills/code-sandbox.md +187 -187
  10. package/extensions/ai-ml/skills/deep-research.md +146 -146
  11. package/extensions/ai-ml/skills/embedding-search.md +66 -66
  12. package/extensions/ai-ml/skills/fine-tuning-guide.md +74 -74
  13. package/extensions/ai-ml/skills/llm-architect.md +125 -125
  14. package/extensions/ai-ml/skills/llm-integration.md +64 -64
  15. package/extensions/ai-ml/skills/prompt-patterns.md +72 -72
  16. package/extensions/ai-ml/skills/rag-patterns.md +66 -66
  17. package/extensions/ai-ml/skills/web-extraction.md +114 -114
  18. package/extensions/analytics/PACK.md +92 -92
  19. package/extensions/analytics/skills/ab-testing.md +72 -72
  20. package/extensions/analytics/skills/dashboard-patterns.md +83 -83
  21. package/extensions/analytics/skills/data-validation.md +68 -68
  22. package/extensions/analytics/skills/funnel-analysis.md +81 -81
  23. package/extensions/analytics/skills/sql-patterns.md +57 -57
  24. package/extensions/analytics/skills/statistical-analysis.md +79 -79
  25. package/extensions/analytics/skills/tracking-setup.md +71 -71
  26. package/extensions/backend/PACK.md +104 -104
  27. package/extensions/backend/skills/api-patterns.md +84 -84
  28. package/extensions/backend/skills/async-pipeline.md +193 -193
  29. package/extensions/backend/skills/auth-patterns.md +97 -97
  30. package/extensions/backend/skills/background-jobs.md +133 -133
  31. package/extensions/backend/skills/caching-patterns.md +108 -108
  32. package/extensions/backend/skills/cli-generation.md +133 -133
  33. package/extensions/backend/skills/database-patterns.md +87 -87
  34. package/extensions/backend/skills/middleware-patterns.md +104 -104
  35. package/extensions/chrome-ext/PACK.md +93 -93
  36. package/extensions/chrome-ext/skills/cws-preflight.md +143 -143
  37. package/extensions/chrome-ext/skills/cws-publish.md +104 -104
  38. package/extensions/chrome-ext/skills/ext-ai-integration.md +251 -251
  39. package/extensions/chrome-ext/skills/ext-messaging.md +139 -139
  40. package/extensions/chrome-ext/skills/ext-storage.md +133 -133
  41. package/extensions/chrome-ext/skills/mv3-scaffold.md +164 -164
  42. package/extensions/content/PACK.md +96 -96
  43. package/extensions/content/skills/blog-patterns.md +88 -88
  44. package/extensions/content/skills/cms-integration.md +131 -131
  45. package/extensions/content/skills/content-scoring.md +107 -107
  46. package/extensions/content/skills/i18n.md +83 -83
  47. package/extensions/content/skills/mdx-authoring.md +137 -137
  48. package/extensions/content/skills/reference.md +1014 -1014
  49. package/extensions/content/skills/seo-patterns.md +67 -67
  50. package/extensions/content/skills/video-repurpose.md +153 -153
  51. package/extensions/devops/PACK.md +101 -101
  52. package/extensions/devops/skills/chaos-testing.md +67 -67
  53. package/extensions/devops/skills/ci-cd.md +75 -75
  54. package/extensions/devops/skills/docker.md +58 -58
  55. package/extensions/devops/skills/edge-serverless.md +163 -163
  56. package/extensions/devops/skills/infra-as-code.md +158 -158
  57. package/extensions/devops/skills/kubernetes.md +110 -110
  58. package/extensions/devops/skills/monitoring.md +57 -57
  59. package/extensions/devops/skills/server-setup.md +64 -64
  60. package/extensions/devops/skills/ssl-domain.md +42 -42
  61. package/extensions/ecommerce/PACK.md +116 -116
  62. package/extensions/ecommerce/skills/cart-system.md +79 -79
  63. package/extensions/ecommerce/skills/inventory-mgmt.md +102 -102
  64. package/extensions/ecommerce/skills/order-management.md +126 -126
  65. package/extensions/ecommerce/skills/payment-integration.md +472 -472
  66. package/extensions/ecommerce/skills/shopify-dev.md +69 -69
  67. package/extensions/ecommerce/skills/subscription-billing.md +93 -93
  68. package/extensions/ecommerce/skills/tax-compliance.md +117 -117
  69. package/extensions/gamedev/PACK.md +142 -142
  70. package/extensions/gamedev/skills/asset-pipeline.md +74 -74
  71. package/extensions/gamedev/skills/audio-system.md +129 -129
  72. package/extensions/gamedev/skills/camera-system.md +87 -87
  73. package/extensions/gamedev/skills/ecs.md +98 -98
  74. package/extensions/gamedev/skills/game-loops.md +72 -72
  75. package/extensions/gamedev/skills/input-system.md +199 -199
  76. package/extensions/gamedev/skills/multiplayer.md +180 -180
  77. package/extensions/gamedev/skills/particles.md +105 -105
  78. package/extensions/gamedev/skills/physics-engine.md +89 -89
  79. package/extensions/gamedev/skills/scene-management.md +146 -146
  80. package/extensions/gamedev/skills/threejs-patterns.md +90 -90
  81. package/extensions/gamedev/skills/webgl.md +71 -71
  82. package/extensions/mobile/PACK.md +106 -106
  83. package/extensions/mobile/skills/app-store-connect.md +152 -152
  84. package/extensions/mobile/skills/app-store-prep.md +66 -66
  85. package/extensions/mobile/skills/deep-linking.md +109 -109
  86. package/extensions/mobile/skills/flutter.md +60 -60
  87. package/extensions/mobile/skills/ios-build-pipeline.md +142 -142
  88. package/extensions/mobile/skills/native-bridge.md +66 -66
  89. package/extensions/mobile/skills/ota-updates.md +97 -97
  90. package/extensions/mobile/skills/push-notifications.md +111 -111
  91. package/extensions/mobile/skills/react-native.md +82 -82
  92. package/extensions/saas/PACK.md +116 -116
  93. package/extensions/saas/skills/billing-integration.md +200 -200
  94. package/extensions/saas/skills/feature-flags.md +130 -130
  95. package/extensions/saas/skills/multi-tenant.md +103 -103
  96. package/extensions/saas/skills/onboarding-flow.md +139 -139
  97. package/extensions/saas/skills/subscription-flow.md +95 -95
  98. package/extensions/saas/skills/team-management.md +144 -144
  99. package/extensions/security/PACK.md +99 -99
  100. package/extensions/security/skills/api-security.md +140 -140
  101. package/extensions/security/skills/compliance.md +68 -68
  102. package/extensions/security/skills/owasp-audit.md +64 -64
  103. package/extensions/security/skills/pentest-patterns.md +77 -77
  104. package/extensions/security/skills/secret-mgmt.md +65 -65
  105. package/extensions/security/skills/supply-chain.md +65 -65
  106. package/extensions/trading/PACK.md +80 -80
  107. package/extensions/trading/skills/chart-components.md +55 -55
  108. package/extensions/trading/skills/experiment-loop.md +125 -125
  109. package/extensions/trading/skills/fintech-patterns.md +47 -47
  110. package/extensions/trading/skills/indicator-library.md +58 -58
  111. package/extensions/trading/skills/quant-analysis.md +111 -111
  112. package/extensions/trading/skills/realtime-data.md +58 -58
  113. package/extensions/trading/skills/trade-logic.md +104 -104
  114. package/extensions/ui/PACK.md +130 -130
  115. package/extensions/ui/skills/a11y-audit.md +91 -91
  116. package/extensions/ui/skills/animation-patterns.md +127 -127
  117. package/extensions/ui/skills/component-patterns.md +100 -100
  118. package/extensions/ui/skills/design-decision.md +108 -108
  119. package/extensions/ui/skills/design-system.md +68 -68
  120. package/extensions/ui/skills/landing-patterns.md +155 -155
  121. package/extensions/ui/skills/palette-picker.md +173 -173
  122. package/extensions/ui/skills/react-health.md +90 -90
  123. package/extensions/ui/skills/type-system.md +125 -125
  124. package/extensions/ui/skills/web-vitals.md +153 -153
  125. package/extensions/zalo/PACK.md +145 -145
  126. package/extensions/zalo/skills/zalo-oa-mcp.md +317 -317
  127. package/extensions/zalo/skills/zalo-oa-messaging.md +429 -429
  128. package/extensions/zalo/skills/zalo-oa-setup.md +236 -236
  129. package/extensions/zalo/skills/zalo-oa-webhook.md +189 -189
  130. package/extensions/zalo/skills/zalo-personal-messaging.md +194 -194
  131. package/extensions/zalo/skills/zalo-personal-setup.md +153 -153
  132. package/extensions/zalo/skills/zalo-rate-guard.md +219 -219
  133. package/hooks/auto-format/index.cjs +48 -48
  134. package/hooks/hooks.json +111 -111
  135. package/hooks/post-session-reflect/index.cjs +189 -189
  136. package/hooks/pre-compact/index.cjs +95 -95
  137. package/hooks/run-hook.cmd +1 -1
  138. package/hooks/secrets-scan/index.cjs +100 -100
  139. package/hooks/session-start/index.cjs +71 -71
  140. package/hooks/typecheck/index.cjs +65 -65
  141. package/package.json +63 -63
  142. package/references/ui-pro-max-data/LICENSE-UI-PRO-MAX +21 -21
  143. package/references/ui-pro-max-data/charts.csv +26 -26
  144. package/references/ui-pro-max-data/colors.csv +161 -161
  145. package/references/ui-pro-max-data/styles.csv +68 -68
  146. package/references/ui-pro-max-data/typography.csv +74 -74
  147. package/references/ui-pro-max-data/ui-reasoning.csv +162 -162
  148. package/references/ui-pro-max-data/ux-guidelines.csv +99 -99
  149. package/skills/adversary/SKILL.md +283 -283
  150. package/skills/asset-creator/SKILL.md +157 -157
  151. package/skills/audit/SKILL.md +147 -2
  152. package/skills/autopsy/SKILL.md +335 -335
  153. package/skills/brainstorm/SKILL.md +342 -342
  154. package/skills/browser-pilot/SKILL.md +168 -168
  155. package/skills/constraint-check/SKILL.md +165 -165
  156. package/skills/context-engine/SKILL.md +404 -404
  157. package/skills/cook/SKILL.md +917 -863
  158. package/skills/db/SKILL.md +273 -273
  159. package/skills/debug/SKILL.md +465 -465
  160. package/skills/dependency-doctor/SKILL.md +265 -235
  161. package/skills/deploy/SKILL.md +274 -231
  162. package/skills/design/DESIGN-REFERENCE.md +365 -365
  163. package/skills/design/SKILL.md +589 -589
  164. package/skills/doc-processor/SKILL.md +254 -254
  165. package/skills/docs/SKILL.md +374 -374
  166. package/skills/docs-seeker/SKILL.md +177 -177
  167. package/skills/fix/SKILL.md +330 -330
  168. package/skills/git/SKILL.md +339 -339
  169. package/skills/hallucination-guard/SKILL.md +219 -219
  170. package/skills/incident/SKILL.md +254 -253
  171. package/skills/integrity-check/SKILL.md +169 -169
  172. package/skills/journal/SKILL.md +240 -240
  173. package/skills/launch/SKILL.md +344 -344
  174. package/skills/logic-guardian/SKILL.md +251 -251
  175. package/skills/marketing/SKILL.md +290 -289
  176. package/skills/mcp-builder/SKILL.md +425 -425
  177. package/skills/neural-memory/SKILL.md +362 -362
  178. package/skills/onboard/SKILL.md +404 -403
  179. package/skills/perf/SKILL.md +346 -346
  180. package/skills/plan/SKILL.md +433 -428
  181. package/skills/preflight/SKILL.md +415 -415
  182. package/skills/problem-solver/SKILL.md +380 -284
  183. package/skills/rescue/SKILL.md +474 -474
  184. package/skills/retro/SKILL.md +3 -1
  185. package/skills/review/SKILL.md +612 -588
  186. package/skills/review-intake/SKILL.md +249 -249
  187. package/skills/safeguard/SKILL.md +200 -200
  188. package/skills/sast/SKILL.md +190 -190
  189. package/skills/scaffold/SKILL.md +328 -287
  190. package/skills/scope-guard/SKILL.md +180 -180
  191. package/skills/scout/SKILL.md +263 -263
  192. package/skills/sentinel/SKILL.md +382 -381
  193. package/skills/sentinel-env/SKILL.md +254 -254
  194. package/skills/sequential-thinking/SKILL.md +234 -234
  195. package/skills/session-bridge/SKILL.md +543 -543
  196. package/skills/skill-forge/SKILL.md +581 -581
  197. package/skills/skill-router/SKILL.md +3 -0
  198. package/skills/surgeon/SKILL.md +215 -215
  199. package/skills/team/SKILL.md +556 -537
  200. package/skills/test/SKILL.md +614 -614
  201. package/skills/trend-scout/SKILL.md +145 -145
  202. package/skills/verification/SKILL.md +326 -326
  203. package/skills/video-creator/SKILL.md +201 -201
  204. package/skills/watchdog/SKILL.md +168 -168
  205. package/skills/worktree/SKILL.md +140 -140
@@ -148,9 +148,12 @@ These 5 skills are the main interface. Most user intents route here first:
148
148
  | Legacy code / rescue / modernize | `rune:rescue` | Old/messy codebase |
149
149
  | Check project health / full audit | `rune:audit` | Quality assessment |
150
150
  | New project / bootstrap / scaffold | `rune:scaffold` | Greenfield project creation |
151
+ | Auto / autopilot / autonomous / "do it all" / "làm hết" / "đi ngủ" | `rune:autopilot` ⚡Pro | Autonomous multi-session execution (requires approved plan + Pro tier installed) |
151
152
 
152
153
  **Default route**: If unclear, route to `rune:cook`. Cook handles 70% of all requests.
153
154
 
155
+ > **Pro skill note**: `rune:autopilot` requires `@rune-pro` installed. If not available, fall back to `rune:cook` with the approved plan and inform user that autopilot is a Pro feature.
156
+
154
157
  #### Tier 2 — Power User Skills (Direct Invocation)
155
158
 
156
159
  For users who know exactly what they want:
@@ -1,215 +1,215 @@
1
- ---
2
- name: surgeon
3
- description: Incremental refactorer. Refactors ONE module per session using proven patterns — Strangler Fig, Branch by Abstraction, Expand-Migrate-Contract.
4
- metadata:
5
- author: runedev
6
- version: "0.2.0"
7
- layer: L2
8
- model: sonnet
9
- group: rescue
10
- tools: "Read, Write, Edit, Bash, Glob, Grep"
11
- ---
12
-
13
- # surgeon
14
-
15
- ## Purpose
16
-
17
- Incremental refactorer that operates on ONE module per session using proven refactoring patterns. Surgeon is precise and safe — it applies small, tested changes with strict blast radius limits. Each surgery session ends with working, tested code committed.
18
-
19
- <HARD-GATE>
20
- - Blast radius MUST be checked before starting (max 5 files)
21
- - Safeguard MUST have run before any edit is made
22
- - Tests MUST pass after every single edit — never accumulate failing tests
23
- - Never refactor two coupled modules in the same session
24
- </HARD-GATE>
25
-
26
- ## Called By (inbound)
27
-
28
- - `rescue` (L1): Phase 2-N SURGERY — one surgery session per module
29
-
30
- ## Calls (outbound)
31
-
32
- - `scout` (L2): understand module dependencies, consumers, and blast radius
33
- - `safeguard` (L2): if untested module found, build safety net first
34
- - `debug` (L2): when refactoring reveals hidden bugs
35
- - `fix` (L2): apply refactoring changes
36
- - `test` (L2): verify after each change
37
- - `review` (L2): quality check on refactored code
38
- - `journal` (L3): update rescue progress
39
-
40
- ## Execution Steps
41
-
42
- ### Step 1 — Pre-surgery scan
43
-
44
- Call `rune:scout` targeting the module to refactor. Ask scout to return:
45
- - All files the module imports (dependencies)
46
- - All files that import the module (consumers)
47
- - Total file count touched (blast radius check)
48
-
49
- ```
50
- Count the unique files that would be modified in this surgery session.
51
- If count > 5 → STOP. Split surgery into smaller sessions.
52
- Report which files are in scope and which must wait for a later session.
53
- ```
54
-
55
- Confirm that `rune:safeguard` has already run for this module (check for `tests/char/<module>.test.ts` and `rune-safeguard-<module>` git tag).
56
-
57
- If safeguard has NOT run, call `rune:safeguard` now before continuing. Do not skip this.
58
-
59
- ### Step 2 — Select refactoring pattern
60
-
61
- Based on module characteristics from scout, choose ONE pattern:
62
-
63
- | Pattern | When to use |
64
- |---|---|
65
- | **Strangler Fig** | Module > 500 LOC with many consumers. New code grows alongside legacy, consumers migrate one by one. |
66
- | **Branch by Abstraction** | Tightly coupled module. Create interface → wrap legacy behind it → build new impl → flip the switch. |
67
- | **Expand-Migrate-Contract** | Changing a function signature or data shape. Expand (add new), migrate callers, contract (remove old). Each phase = one commit. |
68
- | **Extract & Simplify** | Specific function with cyclomatic complexity > 10. Extract sub-functions, simplify conditionals. |
69
-
70
- State the chosen pattern explicitly before starting.
71
-
72
- ### Step 3 — Refactor
73
-
74
- Use `Edit` for all code changes. Rules:
75
- - One logical change per `Edit` call — do not batch unrelated changes
76
- - Changes MUST be small and reversible
77
- - Never rewrite a file from scratch — use targeted edits
78
- - Never change more than 5 files total in this session
79
- - If a change reveals a hidden bug, stop and call `rune:debug` before continuing
80
-
81
- For **Strangler Fig**: Create the new module file first, then update one consumer at a time.
82
-
83
- For **Branch by Abstraction**: Create the interface first (commit), wrap legacy (commit), build new impl (commit), switch (commit). Four commits minimum.
84
-
85
- For **Expand-Migrate-Contract**: Expand (add new API alongside old), migrate each caller (one commit per caller if possible), contract (remove old API last).
86
-
87
- For **Extract & Simplify**: Extract sub-functions one at a time. Each extraction = one commit.
88
-
89
- ### Step 4 — Test after each change
90
-
91
- After every `Edit`, call `rune:test` targeting:
92
- 1. The characterization tests from `tests/char/<module>.test.ts`
93
- 2. Any existing unit tests for the module
94
- 3. Any consumer tests affected by this change
95
-
96
- ```
97
- If any test fails → STOP. Do NOT continue with more edits.
98
- Call rune:debug to investigate. Fix before next edit.
99
- The code MUST stay in a working state after every single change.
100
- ```
101
-
102
- ### Step 5 — Review
103
-
104
- After all edits for this session are complete and tests pass, call `rune:review` on the changed files.
105
-
106
- Address any CRITICAL or HIGH issues raised by review before committing.
107
-
108
- ### Step 6 — Commit
109
-
110
- Use `Bash` to commit this surgery step:
111
-
112
- ```bash
113
- git add <changed files>
114
- git commit -m "refactor(<module>): [pattern] — [what was done]"
115
- ```
116
-
117
- The commit message MUST describe which pattern was used and what changed. Each commit must leave the codebase in a fully working state.
118
-
119
- ### Step 7 — Update journal
120
-
121
- Call `rune:journal` to record:
122
- - Module operated on
123
- - Pattern used
124
- - Files changed
125
- - Health score delta (estimated)
126
- - What remains for next session (if partial)
127
-
128
- ## Refactoring Patterns
129
-
130
- ```
131
- STRANGLER FIG — New code grows around legacy (module > 500 LOC, many consumers)
132
- BRANCH BY ABSTRACTION — Interface → wrap legacy → build new → switch
133
- EXPAND-MIGRATE-CONTRACT — Each step is one safe commit
134
- EXTRACT & SIMPLIFY — For complex functions (cyclomatic > 10)
135
- ```
136
-
137
- ## Safety Rules
138
-
139
- ```
140
- - NEVER refactor 2 coupled modules in same session
141
- - ALWAYS run tests after each change
142
- - Max blast radius: 5 files per session
143
- - If context low → STOP, save state, commit partial work
144
- - Each commit must leave code in working state
145
- - Never skip safeguard, even for "simple" changes
146
- ```
147
-
148
- ## Output Format
149
-
150
- ```
151
- ## Surgery Report: [Module Name]
152
- - **Pattern**: [chosen pattern]
153
- - **Status**: complete | partial (safe stopping point reached)
154
- - **Health**: [before] → [after estimated]
155
- - **Files Changed**: [list, max 5]
156
- - **Commits**: [count]
157
-
158
- ### Steps Taken
159
- 1. [step] — [result] — [test status]
160
-
161
- ### Remaining (if partial)
162
- - [what's left for next surgery session]
163
- - Recommended: re-run rune:surgeon targeting [module] — session 2
164
-
165
- ### Next Step
166
- [if complete]: Run rune:autopsy to update health scores
167
- [if partial]: Commit this checkpoint, then start new surgeon session for remaining work
168
- ```
169
-
170
- ## Constraints
171
-
172
- 1. MUST verify safeguard tests pass before making any edit
173
- 2. MUST check blast radius before starting — max 5 files per session
174
- 3. MUST run tests after EVERY individual edit — never accumulate untested changes
175
- 4. MUST NOT change function signatures without updating all callers
176
- 5. MUST preserve external behavior — refactoring changes structure, not behavior
177
-
178
- ## Sharp Edges
179
-
180
- Known failure modes for this skill. Check these before declaring done.
181
-
182
- | Failure Mode | Severity | Mitigation |
183
- |---|---|---|
184
- | Editing without confirming safeguard ran first | CRITICAL | HARD-GATE: check for `tests/char/<module>.test.*` AND `rune-safeguard-<module>` tag before first edit |
185
- | Exceeding 5-file blast radius without splitting | HIGH | HARD-GATE: count files in scope before starting — stop and split if > 5 |
186
- | Batching multiple edits before running tests | HIGH | HARD-GATE: run tests after every single Edit call — never accumulate untested changes |
187
- | Wrong pattern chosen for module size/type | MEDIUM | Match pattern explicitly: Strangler Fig = large/many-consumers, Extract = high cyclomatic complexity |
188
- | Not committing at safe stopping points when context runs low | MEDIUM | Every commit = working state — stop before context limit, not after losing partial work |
189
-
190
- ## Done When
191
-
192
- - Safeguard confirmed (char tests + rollback tag exist)
193
- - Blast radius checked and within 5 files
194
- - Refactoring pattern selected and stated explicitly
195
- - All edits applied with tests passing after each individual edit
196
- - Characterization tests still pass after all changes
197
- - review passed on changed files
198
- - Surgery committed with message format `refactor(<module>): <pattern> — <description>`
199
- - journal updated with module health delta and remaining work
200
-
201
- ## Returns
202
-
203
- | Artifact | Format | Location |
204
- |----------|--------|----------|
205
- | Refactored module | Edited source files (max 5) | in-place |
206
- | Before/after diff | Git diff | via `git diff` |
207
- | Surgery Report | Markdown | inline |
208
- | Git commit(s) | Conventional commits | git history |
209
- | Journal entry | Text | via `journal` L3 |
210
-
211
- ## Cost Profile
212
-
213
- ~3000-6000 tokens input, ~1000-2000 tokens output. Sonnet. One module per session.
214
-
215
- **Scope guardrail:** surgeon operates on ONE module per session (max 5 files). Any work beyond that scope must be deferred to a separate surgeon session.
1
+ ---
2
+ name: surgeon
3
+ description: Incremental refactorer. Refactors ONE module per session using proven patterns — Strangler Fig, Branch by Abstraction, Expand-Migrate-Contract.
4
+ metadata:
5
+ author: runedev
6
+ version: "0.2.0"
7
+ layer: L2
8
+ model: sonnet
9
+ group: rescue
10
+ tools: "Read, Write, Edit, Bash, Glob, Grep"
11
+ ---
12
+
13
+ # surgeon
14
+
15
+ ## Purpose
16
+
17
+ Incremental refactorer that operates on ONE module per session using proven refactoring patterns. Surgeon is precise and safe — it applies small, tested changes with strict blast radius limits. Each surgery session ends with working, tested code committed.
18
+
19
+ <HARD-GATE>
20
+ - Blast radius MUST be checked before starting (max 5 files)
21
+ - Safeguard MUST have run before any edit is made
22
+ - Tests MUST pass after every single edit — never accumulate failing tests
23
+ - Never refactor two coupled modules in the same session
24
+ </HARD-GATE>
25
+
26
+ ## Called By (inbound)
27
+
28
+ - `rescue` (L1): Phase 2-N SURGERY — one surgery session per module
29
+
30
+ ## Calls (outbound)
31
+
32
+ - `scout` (L2): understand module dependencies, consumers, and blast radius
33
+ - `safeguard` (L2): if untested module found, build safety net first
34
+ - `debug` (L2): when refactoring reveals hidden bugs
35
+ - `fix` (L2): apply refactoring changes
36
+ - `test` (L2): verify after each change
37
+ - `review` (L2): quality check on refactored code
38
+ - `journal` (L3): update rescue progress
39
+
40
+ ## Execution Steps
41
+
42
+ ### Step 1 — Pre-surgery scan
43
+
44
+ Call `rune:scout` targeting the module to refactor. Ask scout to return:
45
+ - All files the module imports (dependencies)
46
+ - All files that import the module (consumers)
47
+ - Total file count touched (blast radius check)
48
+
49
+ ```
50
+ Count the unique files that would be modified in this surgery session.
51
+ If count > 5 → STOP. Split surgery into smaller sessions.
52
+ Report which files are in scope and which must wait for a later session.
53
+ ```
54
+
55
+ Confirm that `rune:safeguard` has already run for this module (check for `tests/char/<module>.test.ts` and `rune-safeguard-<module>` git tag).
56
+
57
+ If safeguard has NOT run, call `rune:safeguard` now before continuing. Do not skip this.
58
+
59
+ ### Step 2 — Select refactoring pattern
60
+
61
+ Based on module characteristics from scout, choose ONE pattern:
62
+
63
+ | Pattern | When to use |
64
+ |---|---|
65
+ | **Strangler Fig** | Module > 500 LOC with many consumers. New code grows alongside legacy, consumers migrate one by one. |
66
+ | **Branch by Abstraction** | Tightly coupled module. Create interface → wrap legacy behind it → build new impl → flip the switch. |
67
+ | **Expand-Migrate-Contract** | Changing a function signature or data shape. Expand (add new), migrate callers, contract (remove old). Each phase = one commit. |
68
+ | **Extract & Simplify** | Specific function with cyclomatic complexity > 10. Extract sub-functions, simplify conditionals. |
69
+
70
+ State the chosen pattern explicitly before starting.
71
+
72
+ ### Step 3 — Refactor
73
+
74
+ Use `Edit` for all code changes. Rules:
75
+ - One logical change per `Edit` call — do not batch unrelated changes
76
+ - Changes MUST be small and reversible
77
+ - Never rewrite a file from scratch — use targeted edits
78
+ - Never change more than 5 files total in this session
79
+ - If a change reveals a hidden bug, stop and call `rune:debug` before continuing
80
+
81
+ For **Strangler Fig**: Create the new module file first, then update one consumer at a time.
82
+
83
+ For **Branch by Abstraction**: Create the interface first (commit), wrap legacy (commit), build new impl (commit), switch (commit). Four commits minimum.
84
+
85
+ For **Expand-Migrate-Contract**: Expand (add new API alongside old), migrate each caller (one commit per caller if possible), contract (remove old API last).
86
+
87
+ For **Extract & Simplify**: Extract sub-functions one at a time. Each extraction = one commit.
88
+
89
+ ### Step 4 — Test after each change
90
+
91
+ After every `Edit`, call `rune:test` targeting:
92
+ 1. The characterization tests from `tests/char/<module>.test.ts`
93
+ 2. Any existing unit tests for the module
94
+ 3. Any consumer tests affected by this change
95
+
96
+ ```
97
+ If any test fails → STOP. Do NOT continue with more edits.
98
+ Call rune:debug to investigate. Fix before next edit.
99
+ The code MUST stay in a working state after every single change.
100
+ ```
101
+
102
+ ### Step 5 — Review
103
+
104
+ After all edits for this session are complete and tests pass, call `rune:review` on the changed files.
105
+
106
+ Address any CRITICAL or HIGH issues raised by review before committing.
107
+
108
+ ### Step 6 — Commit
109
+
110
+ Use `Bash` to commit this surgery step:
111
+
112
+ ```bash
113
+ git add <changed files>
114
+ git commit -m "refactor(<module>): [pattern] — [what was done]"
115
+ ```
116
+
117
+ The commit message MUST describe which pattern was used and what changed. Each commit must leave the codebase in a fully working state.
118
+
119
+ ### Step 7 — Update journal
120
+
121
+ Call `rune:journal` to record:
122
+ - Module operated on
123
+ - Pattern used
124
+ - Files changed
125
+ - Health score delta (estimated)
126
+ - What remains for next session (if partial)
127
+
128
+ ## Refactoring Patterns
129
+
130
+ ```
131
+ STRANGLER FIG — New code grows around legacy (module > 500 LOC, many consumers)
132
+ BRANCH BY ABSTRACTION — Interface → wrap legacy → build new → switch
133
+ EXPAND-MIGRATE-CONTRACT — Each step is one safe commit
134
+ EXTRACT & SIMPLIFY — For complex functions (cyclomatic > 10)
135
+ ```
136
+
137
+ ## Safety Rules
138
+
139
+ ```
140
+ - NEVER refactor 2 coupled modules in same session
141
+ - ALWAYS run tests after each change
142
+ - Max blast radius: 5 files per session
143
+ - If context low → STOP, save state, commit partial work
144
+ - Each commit must leave code in working state
145
+ - Never skip safeguard, even for "simple" changes
146
+ ```
147
+
148
+ ## Output Format
149
+
150
+ ```
151
+ ## Surgery Report: [Module Name]
152
+ - **Pattern**: [chosen pattern]
153
+ - **Status**: complete | partial (safe stopping point reached)
154
+ - **Health**: [before] → [after estimated]
155
+ - **Files Changed**: [list, max 5]
156
+ - **Commits**: [count]
157
+
158
+ ### Steps Taken
159
+ 1. [step] — [result] — [test status]
160
+
161
+ ### Remaining (if partial)
162
+ - [what's left for next surgery session]
163
+ - Recommended: re-run rune:surgeon targeting [module] — session 2
164
+
165
+ ### Next Step
166
+ [if complete]: Run rune:autopsy to update health scores
167
+ [if partial]: Commit this checkpoint, then start new surgeon session for remaining work
168
+ ```
169
+
170
+ ## Constraints
171
+
172
+ 1. MUST verify safeguard tests pass before making any edit
173
+ 2. MUST check blast radius before starting — max 5 files per session
174
+ 3. MUST run tests after EVERY individual edit — never accumulate untested changes
175
+ 4. MUST NOT change function signatures without updating all callers
176
+ 5. MUST preserve external behavior — refactoring changes structure, not behavior
177
+
178
+ ## Sharp Edges
179
+
180
+ Known failure modes for this skill. Check these before declaring done.
181
+
182
+ | Failure Mode | Severity | Mitigation |
183
+ |---|---|---|
184
+ | Editing without confirming safeguard ran first | CRITICAL | HARD-GATE: check for `tests/char/<module>.test.*` AND `rune-safeguard-<module>` tag before first edit |
185
+ | Exceeding 5-file blast radius without splitting | HIGH | HARD-GATE: count files in scope before starting — stop and split if > 5 |
186
+ | Batching multiple edits before running tests | HIGH | HARD-GATE: run tests after every single Edit call — never accumulate untested changes |
187
+ | Wrong pattern chosen for module size/type | MEDIUM | Match pattern explicitly: Strangler Fig = large/many-consumers, Extract = high cyclomatic complexity |
188
+ | Not committing at safe stopping points when context runs low | MEDIUM | Every commit = working state — stop before context limit, not after losing partial work |
189
+
190
+ ## Done When
191
+
192
+ - Safeguard confirmed (char tests + rollback tag exist)
193
+ - Blast radius checked and within 5 files
194
+ - Refactoring pattern selected and stated explicitly
195
+ - All edits applied with tests passing after each individual edit
196
+ - Characterization tests still pass after all changes
197
+ - review passed on changed files
198
+ - Surgery committed with message format `refactor(<module>): <pattern> — <description>`
199
+ - journal updated with module health delta and remaining work
200
+
201
+ ## Returns
202
+
203
+ | Artifact | Format | Location |
204
+ |----------|--------|----------|
205
+ | Refactored module | Edited source files (max 5) | in-place |
206
+ | Before/after diff | Git diff | via `git diff` |
207
+ | Surgery Report | Markdown | inline |
208
+ | Git commit(s) | Conventional commits | git history |
209
+ | Journal entry | Text | via `journal` L3 |
210
+
211
+ ## Cost Profile
212
+
213
+ ~3000-6000 tokens input, ~1000-2000 tokens output. Sonnet. One module per session.
214
+
215
+ **Scope guardrail:** surgeon operates on ONE module per session (max 5 files). Any work beyond that scope must be deferred to a separate surgeon session.