rbin-task-flow 1.19.4 → 1.23.0
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/.claude/skills/rbin-coding-standards/SKILL.md +29 -0
- package/.claude/skills/rbin-coding-standards/reference.md +42 -0
- package/.claude/skills/rbin-git/SKILL.md +39 -0
- package/.claude/skills/task-flow-audit/SKILL.md +15 -0
- package/.claude/skills/task-flow-check/SKILL.md +15 -0
- package/.claude/skills/task-flow-estimate/SKILL.md +15 -0
- package/.claude/skills/task-flow-generate-flow/SKILL.md +15 -0
- package/.claude/skills/task-flow-improve-changes/SKILL.md +15 -0
- package/.claude/skills/task-flow-refactor/SKILL.md +14 -0
- package/.claude/skills/task-flow-report/SKILL.md +15 -0
- package/.claude/skills/task-flow-review/SKILL.md +14 -0
- package/.claude/skills/task-flow-run/SKILL.md +28 -0
- package/.claude/skills/task-flow-run/workflow.md +59 -0
- package/.claude/skills/task-flow-status/SKILL.md +13 -0
- package/.claude/skills/task-flow-sync/SKILL.md +30 -0
- package/.claude/skills/task-flow-sync/workflow.md +57 -0
- package/.claude/skills/task-flow-think/SKILL.md +17 -0
- package/.codex/config.toml +10 -0
- package/.cursor/rules/code_comments.mdc +4 -4
- package/.cursor/rules/coding_standards.mdc +57 -810
- package/.cursor/rules/commit_practices.mdc +5 -138
- package/.cursor/rules/cursor_rules.mdc +4 -3
- package/.cursor/rules/git_control.mdc +5 -86
- package/.cursor/rules/graphify-task-flow.mdc +31 -0
- package/.cursor/rules/rbin-git-policy.mdc +47 -0
- package/.cursor/rules/self_improve.mdc +3 -3
- package/.cursor/rules/task-flow-cursor.mdc +51 -0
- package/.cursor/rules/task-flow-sync.mdc +46 -0
- package/.cursor/rules/task_analysis.mdc +31 -179
- package/.cursor/rules/task_audit.mdc +6 -5
- package/.cursor/rules/task_check.mdc +2 -3
- package/.cursor/rules/task_estimate.mdc +3 -4
- package/.cursor/rules/task_execution.mdc +26 -138
- package/.cursor/rules/task_generate_flow.mdc +2 -3
- package/.cursor/rules/task_generation.mdc +22 -140
- package/.cursor/rules/task_improve_changes.mdc +3 -4
- package/.cursor/rules/task_refactor.mdc +4 -5
- package/.cursor/rules/task_report.mdc +3 -4
- package/.cursor/rules/task_review.mdc +4 -5
- package/.cursor/rules/task_status.mdc +4 -4
- package/.cursor/rules/task_work.mdc +23 -210
- package/.task-flow/AI-PLATFORMS.md +104 -0
- package/.task-flow/CODEX.md +141 -0
- package/.task-flow/CURSOR.md +94 -0
- package/.task-flow/GRAPHIFY.md +112 -0
- package/.task-flow/OPTIMIZATION-IMPLEMENTATION-TASKS.md +365 -0
- package/.task-flow/OPTIMIZATION-PLAN.md +264 -0
- package/.task-flow/README.md +19 -4
- package/.task-flow/docs/coding-standards-full.md +851 -0
- package/.task-flow/platforms/claude-code.md +352 -0
- package/.task-flow/platforms/codex.md +379 -0
- package/.task-flow/platforms/cursor.md +333 -0
- package/AGENTS.md +69 -31
- package/CLAUDE.md +56 -48
- package/README.md +86 -10
- package/bin/cli.js +41 -16
- package/lib/codex.js +45 -0
- package/lib/cursor.js +41 -0
- package/lib/gitignore.js +101 -0
- package/lib/graphify.js +118 -0
- package/lib/install.js +106 -52
- package/lib/profiles.js +110 -0
- package/lib/skills.js +34 -0
- package/lib/utils.js +38 -2
- package/package.json +6 -2
|
@@ -0,0 +1,264 @@
|
|
|
1
|
+
# Plano de melhorias — RBIN Task Flow (otimização e tokens)
|
|
2
|
+
|
|
3
|
+
Documento de referência com oportunidades identificadas na auditoria do pacote (v1.22+). Objetivo: **Task Flow não deve consumir mais contexto do que um projeto sem ele**, exceto quando o usuário pede workflows pesados de forma explícita.
|
|
4
|
+
|
|
5
|
+
**Relacionado:** [AI-PLATFORMS.md](AI-PLATFORMS.md) · [CURSOR.md](CURSOR.md) · [GRAPHIFY.md](GRAPHIFY.md) · [platforms/cursor.md](platforms/cursor.md)
|
|
6
|
+
|
|
7
|
+
**Implementação (subtarefas para IA):** [OPTIMIZATION-IMPLEMENTATION-TASKS.md](OPTIMIZATION-IMPLEMENTATION-TASKS.md)
|
|
8
|
+
|
|
9
|
+
---
|
|
10
|
+
|
|
11
|
+
## 1. Veredito executivo
|
|
12
|
+
|
|
13
|
+
| Cenário | Situação atual (v1.22) | Meta |
|
|
14
|
+
|---------|------------------------|------|
|
|
15
|
+
| Chat genérico | ~2 regras always-on (~1,5–2k tokens após P0.1) | ≤ ~1,5–2k always-on |
|
|
16
|
+
| `task-flow: run` com `src/` no contexto | Pico **12–18k+** tokens só em regras | ≤ ~3k baseline + skill sob demanda |
|
|
17
|
+
| `task-flow: audit` / implementação Codex | Carrega `coding_standards.mdc` inteiro (~7,5k tokens) | Checklist ~1k; full doc só sob pedido |
|
|
18
|
+
| Claude `/task-flow-run` | Skill + `workflow.md` ~600–800 tokens | Manter |
|
|
19
|
+
| Codex | `AGENTS.md` ~2k sempre | Manter |
|
|
20
|
+
|
|
21
|
+
**Princípio:** dados leves em `.task-flow/` + procedimentos **sob demanda** (skills, `@`, leitura pontual). Anything always-on > ~1,5k tokens precisa justificativa forte.
|
|
22
|
+
|
|
23
|
+
---
|
|
24
|
+
|
|
25
|
+
## 2. Baseline medido (repositório template)
|
|
26
|
+
|
|
27
|
+
Estimativa ~4 caracteres/token.
|
|
28
|
+
|
|
29
|
+
| Camada | Tamanho | Tokens (~) | Quando entra |
|
|
30
|
+
|--------|---------|------------|--------------|
|
|
31
|
+
| `task-flow-cursor.mdc` | ~2 KB | ~0,5k | Sempre (Cursor) |
|
|
32
|
+
| `rbin-git-policy.mdc` | ~1,5 KB | ~0,4k | Sempre (P0.1) |
|
|
33
|
+
| `git_control.mdc` / `commit_practices.mdc` | legado | — | Manual / `@` |
|
|
34
|
+
| **Subtotal always-on** | **~3,5 KB** | **~0,9k** | `task-flow-cursor` + `rbin-git-policy` |
|
|
35
|
+
| `coding_standards.mdc` (checklist) | ~4,4 KB (~100 linhas) | **~1,1k** | Glob `src/**`, `app/**` |
|
|
36
|
+
| `docs/coding-standards-full.md` | sob demanda | — | Audit profundo / seções |
|
|
37
|
+
| `task_work.mdc` | ~40 linhas (stub) | ~0,4k | Apply Intelligently — use `@task-flow-run` |
|
|
38
|
+
| `task_execution.mdc` | stub ~28 linhas | ~0,3k | Intelligent — prefer `task-flow-cursor` |
|
|
39
|
+
| `task-flow-sync.mdc` | ~46 linhas | ~0,5k | Glob `.task-flow/**` — sync |
|
|
40
|
+
| `task_generation.mdc` | ~37 linhas | ~0,4k | Glob — subtask templates |
|
|
41
|
+
| `task_analysis.mdc` | ~43 linhas | ~0,4k | `think` only |
|
|
42
|
+
| 14 skills (total) | ~11 KB | sob invocação | `@` ou `/` |
|
|
43
|
+
| Todas as `.mdc` | ~122 KB | até ~30k | Pior caso (matches errados) |
|
|
44
|
+
|
|
45
|
+
---
|
|
46
|
+
|
|
47
|
+
## 3. O que já está bem (não regredir)
|
|
48
|
+
|
|
49
|
+
- [x] Bootstrap `task-flow-cursor.mdc` compacto com tabela de skills
|
|
50
|
+
- [x] 2 regras `alwaysApply: true` — `task-flow-cursor`, `rbin-git-policy` (vs ~19 antes)
|
|
51
|
+
- [x] Skills com `disable-model-invocation: true` nos workflows (`run`, `sync`, …)
|
|
52
|
+
- [x] `paths: [".task-flow/**"]` em skills de task
|
|
53
|
+
- [x] `CLAUDE.md` / `AGENTS.md` como índice, não manual completo
|
|
54
|
+
- [x] Graphify coexistência sem `graphify.mdc` always-on upstream
|
|
55
|
+
- [x] Codex não embute `coding_standards` no `AGENTS.md`
|
|
56
|
+
- [x] `reset --graphify` para reaplicar template + grafo
|
|
57
|
+
|
|
58
|
+
---
|
|
59
|
+
|
|
60
|
+
## 4. Problemas identificados
|
|
61
|
+
|
|
62
|
+
### 4.1 Tripla cobertura de git — ✅ resolvido (P0.1)
|
|
63
|
+
|
|
64
|
+
| Fonte | Papel |
|
|
65
|
+
|-------|--------|
|
|
66
|
+
| `rbin-git-policy.mdc` | always (único) |
|
|
67
|
+
| `git_control.mdc` / `commit_practices.mdc` | legado, manual |
|
|
68
|
+
| `task-flow-cursor.mdc` | 2 linhas + `@rbin-git` |
|
|
69
|
+
| `rbin-git` skill | sob demanda (detalhes) |
|
|
70
|
+
|
|
71
|
+
### 4.2 `coding_standards.mdc` — ✅ resolvido (P0.2)
|
|
72
|
+
|
|
73
|
+
- **Checklist** (~100 linhas, glob `src/**`) + **full** em `.task-flow/docs/coding-standards-full.md` sob demanda.
|
|
74
|
+
- Skill `rbin-coding-standards`: `disable-model-invocation: true`; lê seções do full doc só se ambíguo.
|
|
75
|
+
|
|
76
|
+
### 4.3 Duplicação skill + regra no `run` — parcial (P0.3 ✅)
|
|
77
|
+
|
|
78
|
+
- `task_work.mdc` enxugado (~40 linhas); fonte da verdade: `@task-flow-run` + `workflow.md`.
|
|
79
|
+
- Evitar `@task_work` ou `task_execution` + skill juntos — usar só `@task-flow-*`.
|
|
80
|
+
|
|
81
|
+
### 4.4 Duplicação no `sync` — resolvido (P1.1)
|
|
82
|
+
|
|
83
|
+
- **`task-flow-sync.mdc`** — procedimento único de sync
|
|
84
|
+
- `task_generation.mdc` — só templates de subtarefas
|
|
85
|
+
- `task_analysis.mdc` — só `think`
|
|
86
|
+
- Skill `task-flow-sync` + `workflow.md`
|
|
87
|
+
|
|
88
|
+
### 4.5 Meta-regras raramente necessárias
|
|
89
|
+
|
|
90
|
+
- `cursor_rules.mdc`, `self_improve.mdc` — úteis só ao editar o próprio pacote de regras
|
|
91
|
+
|
|
92
|
+
### 4.6 Documentação desatualizada — resolvido (P0.6 / v1.23)
|
|
93
|
+
|
|
94
|
+
- `platforms/cursor.md` e `CURSOR.md` alinhados a 2 always-on + skills-first.
|
|
95
|
+
|
|
96
|
+
### 4.7 Graphify
|
|
97
|
+
|
|
98
|
+
- `GRAPH_REPORT.md` ou saída enorme de `graphify query` no contexto
|
|
99
|
+
- Custo variável por subtarefa
|
|
100
|
+
|
|
101
|
+
### 4.8 Dados Task Flow
|
|
102
|
+
|
|
103
|
+
- Ler `tasks.json` inteiro a cada `run` em projetos grandes
|
|
104
|
+
|
|
105
|
+
### 4.9 Instalação e versionamento
|
|
106
|
+
|
|
107
|
+
- `.cursor/` e `.claude/` no `.gitignore` do cliente — skills não versionadas no repo do app
|
|
108
|
+
- ~~Sem perfil `minimal` vs `standard` no CLI~~ → `init|update|reset --profile minimal|standard` (P2.1)
|
|
109
|
+
|
|
110
|
+
---
|
|
111
|
+
|
|
112
|
+
## 5. Orçamento alvo (tokens de instrução)
|
|
113
|
+
|
|
114
|
+
| Modo | Always-on | Sob demanda típico | Total típico |
|
|
115
|
+
|------|-----------|-------------------|--------------|
|
|
116
|
+
| **Minimal** (meta) | ~1,2k | skill ~0,5k | **~1,7k** |
|
|
117
|
+
| **Run + implementar** | ~1,2k | `@task-flow-run` + checklist standards ~1,5k | **~2,7k** |
|
|
118
|
+
| **Audit completo** | ~1,2k | audit + standards **full** (explícito) | **~10k** (aceitável, raro) |
|
|
119
|
+
|
|
120
|
+
**Situação atual (pior caso):** run + `src/` aberto → **~12k+** só em regras (~4× o alvo de run).
|
|
121
|
+
|
|
122
|
+
---
|
|
123
|
+
|
|
124
|
+
## 6. Plano de implementação
|
|
125
|
+
|
|
126
|
+
### P0 — Quick wins (1 PR, alto ROI)
|
|
127
|
+
|
|
128
|
+
| # | Ação | Arquivos / notas | Status |
|
|
129
|
+
|---|------|------------------|--------|
|
|
130
|
+
| P0.1 | Fundir `git_control` + `commit_practices` em uma regra always (~70–90 linhas) | `.cursor/rules/rbin-git-policy.mdc` ou nome equivalente; remover always duplicado | [x] |
|
|
131
|
+
| P0.2 | Dividir standards: **checklist** (glob) + **full** (sob demanda) | `coding_standards.mdc` enxuto; `.task-flow/docs/coding-standards-full.md` com conteúdo atual | [x] |
|
|
132
|
+
| P0.3 | Enxugar `task_work.mdc` → pointer para `@task-flow-run` / `workflow.md` | ~40 linhas máx. | [x] |
|
|
133
|
+
| P0.4 | Enxugar `task_execution.mdc` → pointer para `task-flow-cursor` + skills | Evitar índice duplicado de `CLAUDE.md` | [x] |
|
|
134
|
+
| P0.5 | `rbin-coding-standards`: `disable-model-invocation: true` | `.claude/skills/` + cópia `.cursor/skills/` | [x] |
|
|
135
|
+
| P0.6 | Atualizar `platforms/cursor.md` (§1, §3, inventário) para v1.23 real | 2 always-on + skills; sem “standards em toda conversa” | [x] |
|
|
136
|
+
| P0.7 | Atualizar referências audit/sync para **checklist**, não full doc | `task_audit.mdc`, skills, `AGENTS.md`, `CODEX.md` | [x] |
|
|
137
|
+
|
|
138
|
+
### P1 — Médio prazo
|
|
139
|
+
|
|
140
|
+
| # | Ação | Status |
|
|
141
|
+
|---|------|--------|
|
|
142
|
+
| P1.1 | Unificar sync: uma regra `task-flow-sync.mdc` (intelligent) OU só skill; `task_analysis` focado em `think` | [x] |
|
|
143
|
+
| P1.2 | `cursor_rules.mdc` + `self_improve.mdc` → **manual only** (`@`) | [x] |
|
|
144
|
+
| P1.3 | Skill `task-flow-run`: “ler só task/subtask ativa em `tasks.json`” | [x] |
|
|
145
|
+
| P1.4 | GRAPHIFY.md: reforçar “não colar GRAPH_REPORT inteiro”; limitar saída de query | [x] |
|
|
146
|
+
| P1.5 | `CURSOR.md` / `CLAUDE.md`: guia “evite @task_work; use @task-flow-run” | [x] |
|
|
147
|
+
| P1.6 | `showNextSteps` / `info`: linha “Always-on ~X KB (~Y tokens)” | [x] |
|
|
148
|
+
|
|
149
|
+
### P2 — Estratégico
|
|
150
|
+
|
|
151
|
+
| # | Ação | Status |
|
|
152
|
+
|---|------|--------|
|
|
153
|
+
| P2.1 | CLI `init|update|reset --profile minimal\|standard` | [x] |
|
|
154
|
+
| P2.2 | CLI `--share-ai-config` (não ignorar `.cursor/skills/` no git do cliente) | [x] |
|
|
155
|
+
| P2.3 | Script `scripts/measure-rule-bytes.js` (regressão de tamanho always-on) | [x] |
|
|
156
|
+
| P2.4 | Entrada em CHANGELOG + bump minor por pacote de P0 | [x] |
|
|
157
|
+
|
|
158
|
+
---
|
|
159
|
+
|
|
160
|
+
## 7. Detalhamento técnico (P0.2 — split coding standards)
|
|
161
|
+
|
|
162
|
+
**Proposta de estrutura:**
|
|
163
|
+
|
|
164
|
+
```text
|
|
165
|
+
.cursor/rules/coding_standards.mdc # ~80–120 linhas: checklist + links
|
|
166
|
+
.task-flow/docs/coding-standards-full.md # conteúdo atual (exemplos, Nest, RHF…)
|
|
167
|
+
.claude/skills/rbin-coding-standards/
|
|
168
|
+
SKILL.md # aponta checklist + “read full doc if ambiguous”
|
|
169
|
+
reference.md # já existe; alinhar paths
|
|
170
|
+
```
|
|
171
|
+
|
|
172
|
+
**Regra de carregamento:**
|
|
173
|
+
|
|
174
|
+
| Comando / situação | Carregar |
|
|
175
|
+
|--------------------|----------|
|
|
176
|
+
| Editar `src/**` | checklist (glob) |
|
|
177
|
+
| `@rbin-coding-standards` | checklist + opcional full |
|
|
178
|
+
| `task-flow: audit` | checklist para score; full só se usuário pedir profundidade |
|
|
179
|
+
| `task-flow: sync` (gerar subtasks) | checklist nas instructions, não full |
|
|
180
|
+
|
|
181
|
+
---
|
|
182
|
+
|
|
183
|
+
## 8. Detalhamento técnico (P0.1 — git unificado)
|
|
184
|
+
|
|
185
|
+
**Conteúdo único always-on:**
|
|
186
|
+
|
|
187
|
+
- Proibição absoluta de git write (lista curta)
|
|
188
|
+
- Git read-only permitido
|
|
189
|
+
- Formato Conventional Commits + Task ID
|
|
190
|
+
- “Após subtarefa done → sugerir commit; usuário executa”
|
|
191
|
+
|
|
192
|
+
**Remover / reduzir:**
|
|
193
|
+
|
|
194
|
+
- `commit_practices.mdc` como always separado → fundido ou `alwaysApply: false` + `@commit_practices`
|
|
195
|
+
- Seção Git longa em `task-flow-cursor.mdc` → uma linha + `@rbin-git`
|
|
196
|
+
|
|
197
|
+
---
|
|
198
|
+
|
|
199
|
+
## 9. Guia para quem usa o pacote (v1.23+)
|
|
200
|
+
|
|
201
|
+
| Faça | Evite |
|
|
202
|
+
|------|--------|
|
|
203
|
+
| `@task-flow-run` para implementar | `@task_work` + “run” sem skill |
|
|
204
|
+
| Status só com `tasks.status.md` / `@task-flow-status` | Abrir muitos arquivos `src/` antes de sync |
|
|
205
|
+
| `@rbin-coding-standards` ao codificar | Depender do glob de standards em chat exploratório |
|
|
206
|
+
| `improve changes` vs audit full | `audit` em monorepo gigante sem necessidade |
|
|
207
|
+
| Graphify: `graphify query "…"` focado | Colar `GRAPH_REPORT.md` no chat |
|
|
208
|
+
|
|
209
|
+
**Atualizar projetos:**
|
|
210
|
+
|
|
211
|
+
```bash
|
|
212
|
+
rbin-task-flow reset --graphify # template + grafo (CLI Graphify no PATH)
|
|
213
|
+
```
|
|
214
|
+
|
|
215
|
+
---
|
|
216
|
+
|
|
217
|
+
## 10. Quando Task Flow economiza vs gasta
|
|
218
|
+
|
|
219
|
+
### Economiza (tempo e retrabalho → menos tokens de exploração)
|
|
220
|
+
|
|
221
|
+
- Instruções por subtarefa em `tasks.json`
|
|
222
|
+
- `tasks.status.md` como fonte de progresso
|
|
223
|
+
- Graphify em `run` (menos grep/arquivos)
|
|
224
|
+
- Git só sugerido (menos correções)
|
|
225
|
+
|
|
226
|
+
### Gasta mais que sem Task Flow
|
|
227
|
+
|
|
228
|
+
- Sessões sem uso de tasks mas com always-on ~2,6k/turno
|
|
229
|
+
- `run` + glob `coding_standards` + `task_work` juntos
|
|
230
|
+
- Audit pedindo standards completos sem escopo
|
|
231
|
+
|
|
232
|
+
---
|
|
233
|
+
|
|
234
|
+
## 11. Critérios de sucesso (Definition of Done)
|
|
235
|
+
|
|
236
|
+
- [x] Always-on total ≤ **~5 KB** (~1,2k tokens) — verificado com `npm run measure:rules` (v1.23.0: ~3,7 KB)
|
|
237
|
+
- [x] `task-flow: run next 1` com um arquivo `src/` no chat: instruções ≤ **~4k tokens** (sem pedir full standards) — skills + checklist glob, não full doc always-on
|
|
238
|
+
- [x] Nenhuma duplicação skill + regra longa para o mesmo comando (run, sync) — `@task-flow-run` / `task-flow-sync.mdc`; stubs curtos
|
|
239
|
+
- [x] Docs `platforms/cursor.md` e `CURSOR.md` alinhados ao comportamento real
|
|
240
|
+
- [x] CHANGELOG documenta breaking changes de regras (v1.23.0 minor)
|
|
241
|
+
|
|
242
|
+
---
|
|
243
|
+
|
|
244
|
+
## 12. Riscos e mitigação
|
|
245
|
+
|
|
246
|
+
| Risco | Mitigação |
|
|
247
|
+
|-------|-----------|
|
|
248
|
+
| Agent deixa de seguir standards sem full doc | Checklist obrigatório + `@rbin-coding-standards` documentado |
|
|
249
|
+
| Apply Intelligently falha em “run” | Manter fallback curto em `task_work.mdc` |
|
|
250
|
+
| Projetos antigos com `commit_practices` always | `update` migra/rebaixa; nota no CHANGELOG |
|
|
251
|
+
| Codex não lê skills | Manter `AGENTS.md` + `CODEX.md` com paths para checklist |
|
|
252
|
+
|
|
253
|
+
---
|
|
254
|
+
|
|
255
|
+
## 13. Histórico do documento
|
|
256
|
+
|
|
257
|
+
| Data | Nota |
|
|
258
|
+
|------|------|
|
|
259
|
+
| 2026-06-02 | Plano inicial pós auditoria v1.22 (Cursor skills, 3 always-on, reset --graphify) |
|
|
260
|
+
| 2026-06-02 | P0–P2 implementado; release npm **v1.23.0** (minor) — 2 always-on, `measure:rules`, profiles, share-ai-config |
|
|
261
|
+
|
|
262
|
+
---
|
|
263
|
+
|
|
264
|
+
*Plano vivo — marcar checkboxes conforme PRs forem mergeados. Não editar manualmente os status das tasks do cliente; este arquivo é meta-documentação do pacote RBIN.*
|
package/.task-flow/README.md
CHANGED
|
@@ -1,5 +1,20 @@
|
|
|
1
1
|
# RBIN Task Flow - Quick Commands
|
|
2
2
|
|
|
3
|
+
**Optimize by AI platform:**
|
|
4
|
+
|
|
5
|
+
| Platform | Guide |
|
|
6
|
+
|----------|--------|
|
|
7
|
+
| Index | [AI-PLATFORMS.md](AI-PLATFORMS.md) |
|
|
8
|
+
| Claude Code | [platforms/claude-code.md](platforms/claude-code.md) |
|
|
9
|
+
| Cursor | [platforms/cursor.md](platforms/cursor.md) |
|
|
10
|
+
| Codex | [platforms/codex.md](platforms/codex.md) |
|
|
11
|
+
| Graphify + Task Flow | [GRAPHIFY.md](GRAPHIFY.md) |
|
|
12
|
+
| Codex | [CODEX.md](CODEX.md) · [platforms/codex.md](platforms/codex.md) |
|
|
13
|
+
| Cursor | [CURSOR.md](CURSOR.md) · [platforms/cursor.md](platforms/cursor.md) |
|
|
14
|
+
| Otimização (tokens / roadmap) | [OPTIMIZATION-PLAN.md](OPTIMIZATION-PLAN.md) |
|
|
15
|
+
| Subtarefas para implementar otimização | [OPTIMIZATION-IMPLEMENTATION-TASKS.md](OPTIMIZATION-IMPLEMENTATION-TASKS.md) |
|
|
16
|
+
| Coding standards (full, on demand) | [docs/coding-standards-full.md](docs/coding-standards-full.md) |
|
|
17
|
+
|
|
3
18
|
## 🚀 Quick Commands
|
|
4
19
|
|
|
5
20
|
| Command | Description |
|
|
@@ -16,9 +31,9 @@
|
|
|
16
31
|
| `task-flow: estimate X` | Estimates time for task X (e.g., `task-flow: estimate 1` or `task-flow: estimate 10,11`) |
|
|
17
32
|
| `task-flow: report X` | Generates implementation report for task X (e.g., `task-flow: report 1` or `task-flow: report 10,11`) |
|
|
18
33
|
| `task-flow: generate flow` | Populates tasks.flow.md with dependencies, estimated hours, and AI model recommendations |
|
|
19
|
-
| `task-flow: audit` | Audits codebase against coding standards
|
|
34
|
+
| `task-flow: audit` | Audits codebase against **coding standards checklist**; full doc on demand |
|
|
20
35
|
| `task-flow: check` | Run lint fix (if available) and build; fix any warnings or errors until both pass |
|
|
21
|
-
| `task-flow: improve changes` | Audit only uncommitted files
|
|
36
|
+
| `task-flow: improve changes` | Audit only uncommitted files vs **checklist** (same as audit, scoped to diff) |
|
|
22
37
|
| `rbin-task-flow audit` | **(CLI)** Lists files with unstaged changes (not yet `git add`) |
|
|
23
38
|
|
|
24
39
|
**See complete details below ↓**
|
|
@@ -42,13 +57,13 @@ Analyzes code and suggests new tasks. Asks before adding to `tasks.input.txt`.
|
|
|
42
57
|
Shows current status of tasks and subtasks from the `tasks.status.md` file.
|
|
43
58
|
|
|
44
59
|
### `task-flow: audit`
|
|
45
|
-
Audits the **entire codebase** against the
|
|
60
|
+
Audits the **entire codebase** against the **checklist** in [coding_standards.mdc](.cursor/rules/coding_standards.mdc). Deep reference: [docs/coding-standards-full.md](docs/coding-standards-full.md) (sections only, on demand). Non-destructive: reports gaps and suggests incremental improvements; the user chooses what to adopt. See [task_audit.mdc](.cursor/rules/task_audit.mdc) for the full flow.
|
|
46
61
|
|
|
47
62
|
### `task-flow: check`
|
|
48
63
|
Runs **lint fix** and **build** for the project. Check `package.json` for a lint-with-fix script (e.g. `lint:fix`, `lint -- --fix`) and a build script; run lint fix first, fix any warnings or errors, then run build and fix until it passes. Use before committing or before `task-flow: improve changes` to ensure the project is clean.
|
|
49
64
|
|
|
50
65
|
### `task-flow: improve changes`
|
|
51
|
-
Same as **task-flow: audit**, but **only for files that were changed and not yet committed** (unstaged + staged). The AI obtains the list via `git diff --name-only HEAD`, and
|
|
66
|
+
Same as **task-flow: audit**, but **only for files that were changed and not yet committed** (unstaged + staged). The AI obtains the list via `git diff --name-only HEAD`, and scores those paths against the **checklist** in `coding_standards.mdc` (not the full standards doc unless depth is needed). Use before committing. Does **not** run lint or build — use `task-flow: check` for that.
|
|
52
67
|
|
|
53
68
|
### `rbin-task-flow audit` (CLI only)
|
|
54
69
|
Lists **unstaged** file paths (modified but not yet `git add`). Run in the project root: `rbin-task-flow audit`. Option: `-p, --path <path>`.
|