Mate11 commited on
Commit
e415e74
Β·
1 Parent(s): 244663c

Borra el openapi.yaml

Browse files
Files changed (3) hide show
  1. .gitignore +4 -0
  2. CLAUDE.md +176 -0
  3. openapi.yaml +0 -0
.gitignore CHANGED
@@ -216,3 +216,7 @@ __marimo__/
216
 
217
  # Streamlit
218
  .streamlit/secrets.toml
 
 
 
 
 
216
 
217
  # Streamlit
218
  .streamlit/secrets.toml
219
+
220
+
221
+ .claude/
222
+ .claude-flow/
CLAUDE.md ADDED
@@ -0,0 +1,176 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Ruflo β€” Claude Code Configuration
2
+
3
+ ## Rules
4
+
5
+ - Do what has been asked; nothing more, nothing less
6
+ - NEVER create files unless absolutely necessary β€” prefer editing existing files
7
+ - NEVER create documentation files unless explicitly requested
8
+ - NEVER save working files or tests to root β€” use `/src`, `/tests`, `/docs`, `/config`, `/scripts`
9
+ - ALWAYS read a file before editing it
10
+ - NEVER commit secrets, credentials, or .env files
11
+ - NEVER add a `Co-Authored-By` trailer to user commits unless this project's `.claude/settings.json` has `attribution.commit` set (#2078). The Claude Code Bash tool may suggest one in its default commit-message template β€” ignore it. `Co-Authored-By` is semantic authorship attribution under git/GitHub convention; the tool is the facilitator, not a co-author.
12
+ - Keep files under 500 lines
13
+ - Validate input at system boundaries
14
+
15
+ ## Agent Comms (SendMessage-First Coordination)
16
+
17
+ Named agents coordinate via `SendMessage`, not polling or shared state.
18
+
19
+ ```
20
+ Lead (you) ←→ architect ←→ developer ←→ tester ←→ reviewer
21
+ (named agents message each other directly)
22
+ ```
23
+
24
+ ### Spawning a Coordinated Team
25
+
26
+ ```javascript
27
+ // ALL agents in ONE message, each knows WHO to message next
28
+ Agent({ prompt: "Research the codebase. SendMessage findings to 'architect'.",
29
+ subagent_type: "researcher", name: "researcher", run_in_background: true })
30
+ Agent({ prompt: "Wait for 'researcher'. Design solution. SendMessage to 'coder'.",
31
+ subagent_type: "system-architect", name: "architect", run_in_background: true })
32
+ Agent({ prompt: "Wait for 'architect'. Implement it. SendMessage to 'tester'.",
33
+ subagent_type: "coder", name: "coder", run_in_background: true })
34
+ Agent({ prompt: "Wait for 'coder'. Write tests. SendMessage results to 'reviewer'.",
35
+ subagent_type: "tester", name: "tester", run_in_background: true })
36
+ Agent({ prompt: "Wait for 'tester'. Review code quality and security.",
37
+ subagent_type: "reviewer", name: "reviewer", run_in_background: true })
38
+
39
+ // Kick off the pipeline
40
+ SendMessage({ to: "researcher", summary: "Start", message: "[task context]" })
41
+ ```
42
+
43
+ ### Patterns
44
+
45
+ | Pattern | Flow | Use When |
46
+ |---------|------|----------|
47
+ | **Pipeline** | A β†’ B β†’ C β†’ D | Sequential dependencies (feature dev) |
48
+ | **Fan-out** | Lead β†’ A, B, C β†’ Lead | Independent parallel work (research) |
49
+ | **Supervisor** | Lead ↔ workers | Ongoing coordination (complex refactor) |
50
+
51
+ ### Rules
52
+
53
+ - ALWAYS name agents β€” `name: "role"` makes them addressable
54
+ - ALWAYS include comms instructions in prompts β€” who to message, what to send
55
+ - Spawn ALL agents in ONE message with `run_in_background: true`
56
+ - After spawning: STOP, tell user what's running, wait for results
57
+ - NEVER poll status β€” agents message back or complete automatically
58
+
59
+ ## Swarm & Routing
60
+
61
+ ### Config
62
+ - **Topology**: hierarchical-mesh (anti-drift)
63
+ - **Max Agents**: 15
64
+ - **Memory**: hybrid
65
+ - **HNSW**: Enabled
66
+ - **Neural**: Enabled
67
+
68
+ ```bash
69
+ npx @claude-flow/cli@latest swarm init --topology hierarchical --max-agents 8 --strategy specialized
70
+ ```
71
+
72
+ ### Agent Routing
73
+
74
+ | Task | Agents | Topology |
75
+ |------|--------|----------|
76
+ | Bug Fix | researcher, coder, tester | hierarchical |
77
+ | Feature | architect, coder, tester, reviewer | hierarchical |
78
+ | Refactor | architect, coder, reviewer | hierarchical |
79
+ | Performance | perf-engineer, coder | hierarchical |
80
+ | Security | security-architect, auditor | hierarchical |
81
+
82
+ ### When to Swarm
83
+ - **YES**: 3+ files, new features, cross-module refactoring, API changes, security, performance
84
+ - **NO**: single file edits, 1-2 line fixes, docs updates, config changes, questions
85
+
86
+ ### 3-Tier Model Routing
87
+
88
+ | Tier | Handler | Use Cases |
89
+ |------|---------|-----------|
90
+ | 1 | Agent Booster (WASM) | Simple transforms β€” skip LLM, use Edit directly |
91
+ | 2 | Haiku | Simple tasks, low complexity |
92
+ | 3 | Sonnet/Opus | Architecture, security, complex reasoning |
93
+
94
+ ## Memory & Learning
95
+
96
+ ### Before Any Task
97
+ ```bash
98
+ npx @claude-flow/cli@latest memory search --query "[task keywords]" --namespace patterns
99
+ npx @claude-flow/cli@latest hooks route --task "[task description]"
100
+ ```
101
+
102
+ ### After Success
103
+ ```bash
104
+ npx @claude-flow/cli@latest memory store --namespace patterns --key "[name]" --value "[what worked]"
105
+ npx @claude-flow/cli@latest hooks post-task --task-id "[id]" --success true --store-results true
106
+ ```
107
+
108
+ ### MCP Tools (use `ToolSearch("keyword")` to discover)
109
+
110
+ | Category | Key Tools |
111
+ |----------|-----------|
112
+ | **Memory** | `memory_store`, `memory_search`, `memory_search_unified` |
113
+ | **Bridge** | `memory_import_claude`, `memory_bridge_status` |
114
+ | **Swarm** | `swarm_init`, `swarm_status`, `swarm_health` |
115
+ | **Agents** | `agent_spawn`, `agent_list`, `agent_status` |
116
+ | **Hooks** | `hooks_route`, `hooks_post-task`, `hooks_worker-dispatch` |
117
+ | **Security** | `aidefence_scan`, `aidefence_is_safe`, `aidefence_has_pii` |
118
+ | **Hive-Mind** | `hive-mind_init`, `hive-mind_consensus`, `hive-mind_spawn` |
119
+
120
+ ### Background Workers
121
+
122
+ | Worker | When |
123
+ |--------|------|
124
+ | `audit` | After security changes |
125
+ | `optimize` | After performance work |
126
+ | `testgaps` | After adding features |
127
+ | `map` | Every 5+ file changes |
128
+ | `document` | After API changes |
129
+
130
+ ```bash
131
+ npx @claude-flow/cli@latest hooks worker dispatch --trigger audit
132
+ ```
133
+
134
+ ## Agents
135
+
136
+ **Core**: `coder`, `reviewer`, `tester`, `planner`, `researcher`
137
+ **Architecture**: `system-architect`, `backend-dev`, `mobile-dev`
138
+ **Security**: `security-architect`, `security-auditor`
139
+ **Performance**: `performance-engineer`, `perf-analyzer`
140
+ **Coordination**: `hierarchical-coordinator`, `mesh-coordinator`, `adaptive-coordinator`
141
+ **GitHub**: `pr-manager`, `code-review-swarm`, `issue-tracker`, `release-manager`
142
+
143
+ Any string works as a custom agent type.
144
+
145
+ ## Build & Test
146
+
147
+ - ALWAYS run tests after code changes
148
+ - ALWAYS verify build succeeds before committing
149
+
150
+ ```bash
151
+ npm run build && npm test
152
+ ```
153
+
154
+ ## CLI Quick Reference
155
+
156
+ ```bash
157
+ npx @claude-flow/cli@latest init --wizard # Setup
158
+ npx @claude-flow/cli@latest swarm init --v3-mode # Start swarm
159
+ npx @claude-flow/cli@latest memory search --query "" # Vector search
160
+ npx @claude-flow/cli@latest hooks route --task "" # Route to agent
161
+ npx @claude-flow/cli@latest doctor --fix # Diagnostics
162
+ npx @claude-flow/cli@latest security scan # Security scan
163
+ npx @claude-flow/cli@latest performance benchmark # Benchmarks
164
+ ```
165
+
166
+ 26 commands, 140+ subcommands. Use `--help` on any command for details.
167
+
168
+ ## Setup
169
+
170
+ ```bash
171
+ claude mcp add claude-flow -- npx -y @claude-flow/cli@latest
172
+ npx @claude-flow/cli@latest daemon start
173
+ npx @claude-flow/cli@latest doctor --fix
174
+ ```
175
+
176
+ **Agent tool** handles execution (agents, files, code, git). **MCP tools** handle coordination (swarm, memory, hooks). **CLI** is the same via Bash.
openapi.yaml DELETED
The diff for this file is too large to render. See raw diff