pi-maestro-flow 0.2.0 → 0.3.0
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.
- package/README.md +44 -72
- package/agents/cross-role-reviewer.md +1 -1
- package/agents/impeccable-agent.md +1 -1
- package/agents/ralph-executor.md +1 -1
- package/agents/team-worker.md +1 -1
- package/agents/workflow-analyzer.md +1 -1
- package/agents/workflow-codebase-mapper.md +1 -1
- package/agents/workflow-collab-planner.md +1 -1
- package/agents/workflow-debugger.md +1 -1
- package/agents/workflow-executor.md +1 -1
- package/agents/workflow-integration-checker.md +1 -1
- package/agents/workflow-nyquist-auditor.md +1 -1
- package/agents/workflow-phase-researcher.md +1 -1
- package/agents/workflow-planner.md +1 -1
- package/agents/workflow-project-researcher.md +1 -1
- package/agents/workflow-reviewer.md +1 -1
- package/agents/workflow-verifier.md +1 -1
- package/package.json +13 -4
- package/skills/codify-to-knowhow/SKILL.md +1 -1
- package/skills/delegation-check/SKILL.md +1 -1
- package/skills/domain-add/SKILL.md +5 -7
- package/skills/insight-challenge/SKILL.md +1 -1
- package/skills/learn-decompose/SKILL.md +4 -4
- package/skills/learn-follow/SKILL.md +5 -5
- package/skills/learn-investigate/SKILL.md +5 -5
- package/skills/learn-second-opinion/SKILL.md +5 -5
- package/skills/maestro/SKILL.md +9 -10
- package/skills/maestro-amend/SKILL.md +3 -3
- package/skills/maestro-analyze/SKILL.md +10 -13
- package/skills/maestro-blueprint/SKILL.md +8 -11
- package/skills/maestro-brainstorm/SKILL.md +11 -14
- package/skills/maestro-collab/SKILL.md +4 -4
- package/skills/maestro-companion/SKILL.md +3 -3
- package/skills/maestro-composer/SKILL.md +11 -12
- package/skills/maestro-execute/SKILL.md +9 -12
- package/skills/maestro-fork/SKILL.md +8 -11
- package/skills/maestro-grill/SKILL.md +13 -16
- package/skills/maestro-guard/SKILL.md +5 -5
- package/skills/maestro-help/SKILL.md +3 -3
- package/skills/maestro-impeccable/SKILL.md +7 -7
- package/skills/maestro-init/SKILL.md +9 -12
- package/skills/maestro-merge/SKILL.md +6 -8
- package/skills/maestro-milestone-audit/SKILL.md +3 -5
- package/skills/maestro-milestone-complete/SKILL.md +9 -12
- package/skills/maestro-milestone-release/SKILL.md +9 -11
- package/skills/maestro-next/SKILL.md +5 -5
- package/skills/maestro-overlay/SKILL.md +16 -16
- package/skills/maestro-plan/SKILL.md +9 -12
- package/skills/maestro-player/SKILL.md +5 -5
- package/skills/maestro-quick/SKILL.md +5 -7
- package/skills/maestro-ralph/SKILL.md +16 -17
- package/skills/maestro-ralph-cli/SKILL.md +6 -6
- package/skills/maestro-ralph-cli-execute/SKILL.md +2 -2
- package/skills/maestro-ralph-execute/SKILL.md +4 -4
- package/skills/maestro-ralph-v2/SKILL.md +18 -19
- package/skills/maestro-roadmap/SKILL.md +8 -9
- package/skills/maestro-swarm-workflow/SKILL.md +9 -9
- package/skills/maestro-tools-execute/SKILL.md +5 -7
- package/skills/maestro-tools-register/SKILL.md +7 -9
- package/skills/maestro-ui-codify/SKILL.md +9 -10
- package/skills/maestro-universal-workflow/SKILL.md +15 -15
- package/skills/maestro-update/SKILL.md +10 -10
- package/skills/manage-codebase-rebuild/SKILL.md +4 -6
- package/skills/manage-drift-realign/SKILL.md +8 -11
- package/skills/manage-harvest/SKILL.md +6 -9
- package/skills/manage-issue/SKILL.md +5 -8
- package/skills/manage-issue-discover/SKILL.md +8 -11
- package/skills/manage-kg-extractors/SKILL.md +3 -3
- package/skills/manage-knowhow/SKILL.md +3 -5
- package/skills/manage-knowhow-capture/SKILL.md +4 -6
- package/skills/manage-knowledge-audit/SKILL.md +6 -9
- package/skills/manage-status/SKILL.md +3 -5
- package/skills/manage-wiki/SKILL.md +5 -7
- package/skills/odyssey-debug/SKILL.md +13 -13
- package/skills/odyssey-improve/SKILL.md +15 -15
- package/skills/odyssey-planex/SKILL.md +16 -16
- package/skills/odyssey-review-test-fix/SKILL.md +11 -11
- package/skills/odyssey-ui/SKILL.md +10 -10
- package/skills/prompt-generator/SKILL.md +5 -5
- package/skills/quality-auto-test/SKILL.md +3 -5
- package/skills/quality-debug/SKILL.md +3 -5
- package/skills/quality-refactor/SKILL.md +3 -5
- package/skills/quality-retrospective/SKILL.md +10 -13
- package/skills/quality-review/SKILL.md +5 -8
- package/skills/quality-sync/SKILL.md +3 -5
- package/skills/quality-test/SKILL.md +4 -6
- package/skills/scholar-anti-ai-writing/SKILL.md +1 -1
- package/skills/scholar-citation-verify/SKILL.md +1 -1
- package/skills/scholar-experiment/SKILL.md +2 -2
- package/skills/scholar-ideation/SKILL.md +7 -7
- package/skills/scholar-latex-organizer/SKILL.md +1 -1
- package/skills/scholar-publish/SKILL.md +3 -3
- package/skills/scholar-rebuttal-pro/SKILL.md +3 -3
- package/skills/scholar-review/SKILL.md +1 -1
- package/skills/scholar-writing/SKILL.md +1 -1
- package/skills/security-audit/SKILL.md +2 -4
- package/skills/skill-generator/SKILL.md +3 -3
- package/skills/skill-iter-tune/SKILL.md +3 -3
- package/skills/skill-simplify/SKILL.md +1 -1
- package/skills/skill-tuning/SKILL.md +1 -1
- package/skills/spec-add/SKILL.md +5 -7
- package/skills/spec-load/SKILL.md +3 -5
- package/skills/spec-remove/SKILL.md +4 -6
- package/skills/spec-setup/SKILL.md +6 -8
- package/skills/team-adversarial-swarm/SKILL.md +5 -5
- package/skills/team-arch-opt/SKILL.md +2 -2
- package/skills/team-brainstorm/SKILL.md +3 -3
- package/skills/team-coordinate/SKILL.md +4 -4
- package/skills/team-designer/SKILL.md +1 -1
- package/skills/team-executor/SKILL.md +3 -3
- package/skills/team-frontend/SKILL.md +2 -2
- package/skills/team-frontend-debug/SKILL.md +3 -3
- package/skills/team-interactive-craft/SKILL.md +2 -2
- package/skills/team-issue/SKILL.md +3 -3
- package/skills/team-lifecycle-v4/SKILL.md +4 -4
- package/skills/team-motion-design/SKILL.md +2 -2
- package/skills/team-perf-opt/SKILL.md +3 -3
- package/skills/team-planex/SKILL.md +2 -2
- package/skills/team-quality-assurance/SKILL.md +3 -3
- package/skills/team-review/SKILL.md +3 -3
- package/skills/team-roadmap-dev/SKILL.md +3 -3
- package/skills/team-swarm/SKILL.md +4 -4
- package/skills/team-tech-debt/SKILL.md +2 -2
- package/skills/team-testing/SKILL.md +3 -3
- package/skills/team-ui-polish/SKILL.md +2 -2
- package/skills/team-uidesign/SKILL.md +2 -2
- package/skills/team-ultra-analyze/SKILL.md +4 -4
- package/skills/team-ux-improve/SKILL.md +3 -3
- package/skills/team-visual-a11y/SKILL.md +2 -2
- package/skills/workflow-skill-designer/SKILL.md +8 -8
- package/templates/business-test-report.json +68 -0
- package/templates/config.json +48 -0
- package/templates/context.md +16 -0
- package/templates/doc-index.json +11 -0
- package/templates/index.json +68 -0
- package/templates/issue.json +36 -0
- package/templates/plan.json +38 -0
- package/templates/project.md +60 -0
- package/templates/reflection-log.md +20 -0
- package/templates/roadmap.md +45 -0
- package/templates/scratch-index.json +19 -0
- package/templates/search-tool.json +1 -0
- package/templates/search-tools.md +47 -0
- package/templates/state.json +18 -0
- package/templates/task-summary.md +21 -0
- package/templates/task.json +89 -0
- package/templates/terminology-template.json +22 -0
- package/templates/uat.md +21 -0
- package/templates/validation.json +31 -0
- package/templates/verification.json +37 -0
- package/templates/workflow.md +14 -0
- package/templates/worktree-scope.json +9 -0
- package/templates/worktrees.json +26 -0
- package/workflows/agy-instructions.md +156 -0
- package/workflows/analyze.md +768 -0
- package/workflows/auto-test.md +705 -0
- package/workflows/blueprint.md +406 -0
- package/workflows/boundary-grill.md +50 -0
- package/workflows/brainstorm.md +508 -0
- package/workflows/business-test.md +574 -0
- package/workflows/chinese-response.md +25 -0
- package/workflows/claude-instructions.md +167 -0
- package/workflows/codebase-rebuild.md +267 -0
- package/workflows/codebase-refresh.md +168 -0
- package/workflows/codex-instructions.md +156 -0
- package/workflows/coding-philosophy.md +70 -0
- package/workflows/command-authoring.md +823 -0
- package/workflows/debug.md +409 -0
- package/workflows/delegate-usage.md +99 -0
- package/workflows/domain-add.md +121 -0
- package/workflows/drift-realign.md +396 -0
- package/workflows/execute.md +631 -0
- package/workflows/explore-usage.md +93 -0
- package/workflows/finish-work.md +124 -0
- package/workflows/fork.md +162 -0
- package/workflows/grill.md +494 -0
- package/workflows/harvest.md +478 -0
- package/workflows/impeccable.md +168 -0
- package/workflows/init.md +157 -0
- package/workflows/instruction-authoring-guide.md +97 -0
- package/workflows/integration-test.md +186 -0
- package/workflows/interview-mechanics.md +8 -0
- package/workflows/issue-analyze.md +20 -0
- package/workflows/issue-discover.md +238 -0
- package/workflows/issue-execute.md +110 -0
- package/workflows/issue-gaps-analyze.codex.md +256 -0
- package/workflows/issue-gaps-analyze.md +215 -0
- package/workflows/issue-plan.md +110 -0
- package/workflows/issue.md +338 -0
- package/workflows/knowhow.md +429 -0
- package/workflows/knowledge-audit.md +355 -0
- package/workflows/learn.md +271 -0
- package/workflows/maestro-chain-execute.md +20 -0
- package/workflows/maestro-link-coordinate.md +114 -0
- package/workflows/maestro-super.md +120 -0
- package/workflows/maestro.md +501 -0
- package/workflows/map.md +74 -0
- package/workflows/merge.md +122 -0
- package/workflows/milestone-audit.md +155 -0
- package/workflows/milestone-complete.md +167 -0
- package/workflows/milestone-release.md +82 -0
- package/workflows/odyssey-base-codex.md +305 -0
- package/workflows/odyssey-base.md +256 -0
- package/workflows/overlays.md +109 -0
- package/workflows/plan.md +520 -0
- package/workflows/quick.md +350 -0
- package/workflows/ralph-amend-goal.md +177 -0
- package/workflows/refactor.md +87 -0
- package/workflows/retrospective.md +471 -0
- package/workflows/review.md +448 -0
- package/workflows/roadmap-common.md +193 -0
- package/workflows/roadmap.md +106 -0
- package/workflows/shell-exec-protocol.md +30 -0
- package/workflows/skill-authoring.md +265 -0
- package/workflows/spec-generate.md +475 -0
- package/workflows/specs-add.md +123 -0
- package/workflows/specs-load.md +76 -0
- package/workflows/specs-remove.md +104 -0
- package/workflows/specs-setup.md +316 -0
- package/workflows/status.md +149 -0
- package/workflows/sync.md +116 -0
- package/workflows/tdd.md +255 -0
- package/workflows/test-gen.md +226 -0
- package/workflows/test.md +385 -0
- package/workflows/tools-spec.md +70 -0
- package/workflows/ui-codify-extract.md +372 -0
- package/workflows/ui-codify-knowhow.md +258 -0
- package/workflows/ui-codify-package.md +159 -0
- package/workflows/ui-codify.md +227 -0
- package/workflows/ui-design.md +393 -0
- package/workflows/ui-style.md +201 -0
- package/workflows/verify.md +557 -0
- package/workflows/wiki-connect.md +153 -0
- package/workflows/wiki-digest.md +180 -0
- package/workflows/wiki-manage.md +113 -0
|
@@ -0,0 +1,156 @@
|
|
|
1
|
+
# Maestro
|
|
2
|
+
|
|
3
|
+
- **Coding Philosophy**: @~/.maestro/workflows/coding-philosophy.md
|
|
4
|
+
|
|
5
|
+
## Delegate & CLI
|
|
6
|
+
|
|
7
|
+
- **CLI Endpoints Config**: @~/.maestro/cli-tools.json
|
|
8
|
+
|
|
9
|
+
`maestro delegate "<PROMPT>" --to <tool> --mode analysis|write` — dispatch tasks to external CLI tools (gemini, codex, claude, opencode).
|
|
10
|
+
Always `run_in_background: true`. Full guide: `cat ~/.maestro/workflows/delegate-usage.md`
|
|
11
|
+
|
|
12
|
+
**Strictly follow the cli-tools.json configuration**
|
|
13
|
+
|
|
14
|
+
## Explore
|
|
15
|
+
|
|
16
|
+
`maestro explore` takes priority over Glob, Grep, and Read. When locating files or searching code patterns, call `maestro explore` first and stop to wait for results.
|
|
17
|
+
|
|
18
|
+
```bash
|
|
19
|
+
maestro explore "FIND: <target + condition>\nSCOPE: <paths>" [more prompts...] [options]
|
|
20
|
+
```
|
|
21
|
+
|
|
22
|
+
Lightweight read-only codebase search. 1 prompt = 1 agent. Not for write-mode/long sessions — use `delegate`.
|
|
23
|
+
|
|
24
|
+
| Option | Description |
|
|
25
|
+
|--------|-------------|
|
|
26
|
+
| `-e, --endpoint <names>` | Endpoint name(s), comma-separated |
|
|
27
|
+
| `--all` | Fan out each prompt to all endpoints |
|
|
28
|
+
| `--max-turns <n>` | Max agent turns per job |
|
|
29
|
+
| `-f, --file <path>` | Load prompts from JSON or text file |
|
|
30
|
+
| `--cd <dir>` | Working directory |
|
|
31
|
+
| `--json` | Output results as JSON |
|
|
32
|
+
|
|
33
|
+
### Context Injection
|
|
34
|
+
|
|
35
|
+
Explore agent 无项目认知,调用前注入上下文:
|
|
36
|
+
|
|
37
|
+
| 注入项 | 写入字段 | 内容 |
|
|
38
|
+
|--------|----------|------|
|
|
39
|
+
| 结构 | SCOPE | 相关目录的具体路径(非通配泛扫) |
|
|
40
|
+
| 领域 | SCOPE | `maestro search` 已返回的关键文件路径 |
|
|
41
|
+
| 约束 | ATTENTION | 框架、语言、命名惯例 |
|
|
42
|
+
|
|
43
|
+
```
|
|
44
|
+
FIND: authentication middleware that validates JWT tokens
|
|
45
|
+
SCOPE: src/middleware/, src/auth/, src/api/routes/
|
|
46
|
+
ATTENTION: Express.js, middleware files named *.middleware.ts
|
|
47
|
+
```
|
|
48
|
+
|
|
49
|
+
### Prompt Structure
|
|
50
|
+
|
|
51
|
+
**FIND + SCOPE 为最低标准。** 每个字段一句陈述句,禁止嵌套条件。
|
|
52
|
+
|
|
53
|
+
| Field | Required | Rule |
|
|
54
|
+
|-------|----------|------|
|
|
55
|
+
| `FIND` | **Yes** | 可判定的具体目标(什么 + 判定条件) |
|
|
56
|
+
| `SCOPE` | **Yes** | 明确路径或 glob,禁止 `**/*` 泛扫 |
|
|
57
|
+
| `EXCLUDE` | No | 要跳过的文件类型或目录 |
|
|
58
|
+
| `ATTENTION` | No | 框架、命名惯例、已知陷阱 |
|
|
59
|
+
| `EXPECTED` | Recommended | 输出格式:`file:line` 列表 / 摘要 / JSON |
|
|
60
|
+
|
|
61
|
+
```
|
|
62
|
+
FIND: Functions that call db.query() with string concatenation instead of $1/$2
|
|
63
|
+
SCOPE: src/db/**/*.ts, src/api/**/*.ts
|
|
64
|
+
EXCLUDE: **/*.test.ts
|
|
65
|
+
EXPECTED: file:line list with the SQL string
|
|
66
|
+
```
|
|
67
|
+
|
|
68
|
+
### Cross-Search
|
|
69
|
+
|
|
70
|
+
对重要搜索,用 2-3 个不同角度的 prompt 并发,结果由 Claude 交叉验证。
|
|
71
|
+
|
|
72
|
+
**按角度拆分,不按关键词拆分:**
|
|
73
|
+
|
|
74
|
+
| 角度 | Prompt A | Prompt B |
|
|
75
|
+
|------|----------|----------|
|
|
76
|
+
| 定义 vs 调用 | 找函数定义 | 找调用点 |
|
|
77
|
+
| 正例 vs 反例 | 找正确用法 | 找遗漏用法 |
|
|
78
|
+
| 入口 vs 实现 | 找 export/路由 | 找内部逻辑 |
|
|
79
|
+
| 按文件类型 | .ts 中的用法 | .vue 中的用法 |
|
|
80
|
+
|
|
81
|
+
```bash
|
|
82
|
+
maestro explore \
|
|
83
|
+
"FIND: All functions exported from auth module\nSCOPE: src/auth/\nEXPECTED: function name + file:line" \
|
|
84
|
+
"FIND: All imports from auth module\nSCOPE: src/**/*.ts\nEXCLUDE: src/auth/\nEXPECTED: import path + file:line" \
|
|
85
|
+
--json
|
|
86
|
+
```
|
|
87
|
+
|
|
88
|
+
**结果置信度:**
|
|
89
|
+
- 双命中 → 高置信,直接使用
|
|
90
|
+
- 单命中 → 用 Grep/Read 二次确认
|
|
91
|
+
- 零命中 → 换角度重搜或目标不存在
|
|
92
|
+
|
|
93
|
+
### Execution
|
|
94
|
+
|
|
95
|
+
Multi-prompt — background;single lookup — foreground:
|
|
96
|
+
|
|
97
|
+
```
|
|
98
|
+
Bash({ command: "maestro explore \"p1\" \"p2\" --json", run_in_background: true })
|
|
99
|
+
Bash({ command: "maestro explore \"FIND: ...\nSCOPE: ...\"" })
|
|
100
|
+
```
|
|
101
|
+
|
|
102
|
+
Session: `maestro explore show` / `maestro explore output <id>`
|
|
103
|
+
|
|
104
|
+
## Knowledge System
|
|
105
|
+
|
|
106
|
+
**Gate rule**: run `maestro search` + `maestro load` BEFORE reading code or editing files.
|
|
107
|
+
|
|
108
|
+
```bash
|
|
109
|
+
maestro search "<query>" [--type <type>] [--category <cat>] [--code] [--kg]
|
|
110
|
+
maestro load --type <type> [--list] [--category <cat>] [--keyword <word>] [--id <id>]
|
|
111
|
+
```
|
|
112
|
+
|
|
113
|
+
**--type**: `spec`, `knowhow`, `domain`, `issue`, `session`, `scratch`, `note`, `project`, `roadmap`
|
|
114
|
+
**--category** (spec only): `coding`, `arch`, `debug`, `test`, `review`, `learning`, `ui`
|
|
115
|
+
|
|
116
|
+
### Query Rules
|
|
117
|
+
|
|
118
|
+
1-3 core keywords per query — multiple short queries beat one long one.
|
|
119
|
+
Separate concepts from symbols. Add `--code` for symbols, `--kg` for full-source.
|
|
120
|
+
|
|
121
|
+
```bash
|
|
122
|
+
# ❌ keyword dump
|
|
123
|
+
maestro search "topology display frontend DetailedTopologySVG elk"
|
|
124
|
+
|
|
125
|
+
# ✅ targeted
|
|
126
|
+
maestro search "topology layout"
|
|
127
|
+
maestro search "DetailedTopologySVG" --code
|
|
128
|
+
maestro load --type spec --category coding
|
|
129
|
+
```
|
|
130
|
+
|
|
131
|
+
### Record
|
|
132
|
+
|
|
133
|
+
| What | Command |
|
|
134
|
+
|------|---------|
|
|
135
|
+
| Spec | `/spec-add <category> "title" "content" --keywords kw1,kw2 --description "summary"` |
|
|
136
|
+
| Knowhow | `/manage-knowhow-capture` (`--spec-category <cat>` for agent injection) |
|
|
137
|
+
|
|
138
|
+
Category routing: decisions→`arch`, patterns→`coding`, pitfalls→`debug`/`learning`, rules→`review`, tests→`test`.
|
|
139
|
+
|
|
140
|
+
### Supersession & Conflict (dual-track)
|
|
141
|
+
|
|
142
|
+
| 关系 | 场景 | 命令 | 效果 |
|
|
143
|
+
|------|------|------|------|
|
|
144
|
+
| **supersede** | 新规则替代旧规则 | `maestro spec supersede <old-sid> --by <new-sid>` | 旧条目 `deprecated`,演化链保留 |
|
|
145
|
+
| **conflict** | 两条规则均有道理 | `maestro spec conflict mark <file> <line> --note "<reason>"` | 旧条目 `contested`(search ×0.5),人裁决 |
|
|
146
|
+
|
|
147
|
+
Confidence levels: `high` → `medium` (default) → `low` (`[LOW CONFIDENCE]`) → `contested` (`[CONTESTED]`).
|
|
148
|
+
Resolution: `/manage-knowledge-audit`
|
|
149
|
+
|
|
150
|
+
### Health & Maintenance
|
|
151
|
+
|
|
152
|
+
```bash
|
|
153
|
+
maestro spec health # 生命周期统计 + 演化链完整性
|
|
154
|
+
maestro spec backfill-sid # 存量无 sid 条目回填
|
|
155
|
+
maestro spec history <sid> # 演化链查看
|
|
156
|
+
```
|
|
@@ -0,0 +1,70 @@
|
|
|
1
|
+
# Coding Philosophy
|
|
2
|
+
|
|
3
|
+
## Core Beliefs
|
|
4
|
+
|
|
5
|
+
- **Pursue good taste** - Eliminate edge cases to make code logic natural and elegant
|
|
6
|
+
- **Embrace extreme simplicity** - Complexity is the root of all evil
|
|
7
|
+
- **Be pragmatic** - Code must solve real-world problems, not hypothetical ones
|
|
8
|
+
- **Data structures first** - Bad programmers worry about code; good programmers worry about data structures
|
|
9
|
+
- **Never break backward compatibility** - Existing functionality is sacred and inviolable
|
|
10
|
+
- **Incremental progress over big bangs** - Small changes that compile and pass tests
|
|
11
|
+
- **Learning from existing code** - Study and plan before implementing
|
|
12
|
+
- **Clear intent over clever code** - Be boring and obvious
|
|
13
|
+
- **Follow existing code style** - Match import patterns, naming conventions, and formatting of existing codebase
|
|
14
|
+
- **Minimize changes** - Only modify what's directly required; avoid refactoring, adding features, or "improving" code beyond the request
|
|
15
|
+
- **No unsolicited documentation** - NEVER generate reports, documentation files, or summaries without explicit user request. If required, save to .workflow/.scratchpad/
|
|
16
|
+
|
|
17
|
+
## Simplicity Means
|
|
18
|
+
|
|
19
|
+
- Single responsibility per function/class
|
|
20
|
+
- Avoid premature abstractions
|
|
21
|
+
- No clever tricks - choose the boring solution
|
|
22
|
+
- If you need to explain it, it's too complex
|
|
23
|
+
|
|
24
|
+
## Fix, Don't Hide
|
|
25
|
+
|
|
26
|
+
**Solve problems, don't silence symptoms** - Skipped tests, `@ts-ignore`, empty catch, `as any`, excessive timeouts = hiding bugs, not fixing them
|
|
27
|
+
|
|
28
|
+
**NEVER**:
|
|
29
|
+
- Make assumptions - verify with existing code
|
|
30
|
+
- Generate reports, summaries, or documentation files without explicit user request
|
|
31
|
+
- Use suppression mechanisms (`skip`, `ignore`, `disable`) without fixing root cause
|
|
32
|
+
|
|
33
|
+
**ALWAYS**:
|
|
34
|
+
- Plan complex tasks thoroughly before implementation
|
|
35
|
+
- Generate task decomposition for multi-module work (>3 modules or >5 subtasks)
|
|
36
|
+
- Track progress using TODO checklists for complex tasks
|
|
37
|
+
- Validate planning documents before starting development
|
|
38
|
+
- Commit working code incrementally
|
|
39
|
+
- Update plan documentation and progress tracking as you go
|
|
40
|
+
- Learn from existing implementations
|
|
41
|
+
- Stop after 3 failed attempts and reassess
|
|
42
|
+
- **Edit fallback**: When Edit tool fails 2+ times on same file, try Bash sed/awk first, then Write to recreate if still failing
|
|
43
|
+
|
|
44
|
+
## Learning the Codebase
|
|
45
|
+
|
|
46
|
+
- Find 3 similar features/components
|
|
47
|
+
- Identify common patterns and conventions
|
|
48
|
+
- Use same libraries/utilities when possible
|
|
49
|
+
- Follow existing test patterns
|
|
50
|
+
|
|
51
|
+
## Tooling
|
|
52
|
+
|
|
53
|
+
- Use project's existing build system
|
|
54
|
+
- Use project's test framework
|
|
55
|
+
- Use project's formatter/linter settings
|
|
56
|
+
- Don't introduce new tools without strong justification
|
|
57
|
+
|
|
58
|
+
## Content Uniqueness Rules
|
|
59
|
+
|
|
60
|
+
- **Each layer owns its abstraction level** - no content sharing between layers
|
|
61
|
+
- **Reference, don't duplicate** - point to other layers, never copy content
|
|
62
|
+
- **Maintain perspective** - each layer sees the system at its appropriate scale
|
|
63
|
+
- **Avoid implementation creep** - higher layers stay architectural
|
|
64
|
+
|
|
65
|
+
# Context Requirements
|
|
66
|
+
|
|
67
|
+
Before implementation, always:
|
|
68
|
+
- Identify 3+ existing similar patterns
|
|
69
|
+
- Map dependencies and integration points
|
|
70
|
+
- Understand testing framework and coding conventions
|