1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
//! Loop depth tracking for nested tool loops.
//!
//! When tools spawn sub-agents (nested `tool_loop` calls), depth tracking
//! prevents runaway recursion. Use [`LoopContext`] for built-in depth
//! management, or implement [`LoopDepth`] manually on your own type.
//!
//! # Depth semantics
//!
//! - **Depth 0** = the outermost (top-level) loop. This is the initial
//! context created by the application before entering any tool loop.
//! - **Depth N** = the context is inside N levels of nested tool loops.
//! - `LoopCore::new()` auto-increments: if you pass a context at depth 0,
//! tools inside that loop receive a context at depth 1. Tools that spawn
//! their own `tool_loop` will pass depth 1 in, and their inner tools
//! see depth 2, etc.
//! - `max_depth` is checked *before* incrementing: if `ctx.loop_depth() >= max_depth`,
//! the loop immediately returns `LlmError::MaxDepthExceeded`. With the default
//! `max_depth = Some(3)`, the deepest allowed nesting is master(0) → worker(1) → sub-worker(2).
//! A tool at depth 2 trying to spawn another loop would be rejected (2 >= 3 is false,
//! but a tool at depth 3 would be: 3 >= 3 is true).
//!
//! # Using `LoopContext` (recommended)
//!
//! ```rust
//! use llm_stack::tool::LoopContext;
//!
//! // Wrap your application state — depth tracking is automatic
//! let ctx = LoopContext::new(MyState { user_id: "u123".into() });
//! # #[derive(Clone)] struct MyState { user_id: String }
//! ```
//!
//! # Manual implementation
//!
//! ```rust
//! use llm_stack::tool::LoopDepth;
//!
//! #[derive(Clone)]
//! struct AgentContext {
//! user_id: String,
//! depth: u32,
//! }
//!
//! impl LoopDepth for AgentContext {
//! fn loop_depth(&self) -> u32 {
//! self.depth
//! }
//!
//! fn with_depth(&self, depth: u32) -> Self {
//! Self {
//! depth,
//! ..self.clone()
//! }
//! }
//! }
//! ```
/// Trait for contexts that support automatic depth tracking in nested tool loops.
///
/// When `tool_loop` executes tools, `LoopCore` calls
/// `ctx.with_depth(ctx.loop_depth() + 1)` and passes the result to tool
/// handlers. If a tool handler then enters its own `tool_loop`, the depth
/// is checked against [`ToolLoopConfig::max_depth`](super::ToolLoopConfig::max_depth)
/// and the loop is rejected if the limit is reached.
///
/// # Blanket Implementation
///
/// The unit type `()` has a blanket implementation that always returns depth 0
/// and ignores `with_depth`. Use this for simple cases where depth tracking
/// isn't needed:
///
/// ```rust
/// use llm_stack::tool::LoopDepth;
///
/// assert_eq!(().loop_depth(), 0);
/// assert_eq!(().with_depth(5), ());
/// ```
///
/// # Custom Implementation
///
/// For agent systems with nesting, implement this on your context type:
///
/// ```rust
/// use llm_stack::tool::LoopDepth;
///
/// #[derive(Clone)]
/// struct MyContext {
/// session_id: String,
/// loop_depth: u32,
/// }
///
/// impl LoopDepth for MyContext {
/// fn loop_depth(&self) -> u32 {
/// self.loop_depth
/// }
///
/// fn with_depth(&self, depth: u32) -> Self {
/// Self {
/// loop_depth: depth,
/// ..self.clone()
/// }
/// }
/// }
/// ```
/// Blanket implementation for unit type — always depth 0, no tracking.
///
/// This allows simple use cases to work without implementing the trait:
///
/// ```rust
/// use llm_stack::tool::{ToolLoopConfig, ToolRegistry};
///
/// // Works with () context, no depth tracking
/// let registry: ToolRegistry<()> = ToolRegistry::new();
/// ```
// ── LoopContext ──────────────────────────────────────────────────────
/// Generic context wrapper with built-in depth tracking.
///
/// Wraps any `Clone + Send + Sync` state and automatically implements
/// [`LoopDepth`], eliminating the boilerplate of storing a `depth` field
/// and writing the trait impl yourself.
///
/// # Examples
///
/// ```rust
/// use llm_stack::tool::{LoopContext, LoopDepth, ToolRegistry};
///
/// #[derive(Clone)]
/// struct AppState {
/// user_id: String,
/// api_key: String,
/// }
///
/// let ctx = LoopContext::new(AppState {
/// user_id: "user_123".into(),
/// api_key: "sk-secret".into(),
/// });
///
/// assert_eq!(ctx.loop_depth(), 0);
/// assert_eq!(ctx.state.user_id, "user_123");
///
/// // Use with a typed registry
/// let registry: ToolRegistry<LoopContext<AppState>> = ToolRegistry::new();
/// ```
///
/// For the zero-state case, use `LoopContext<()>`:
///
/// ```rust
/// use llm_stack::tool::{LoopContext, LoopDepth};
///
/// let ctx = LoopContext::empty();
/// assert_eq!(ctx.loop_depth(), 0);
///
/// let nested = ctx.with_depth(1);
/// assert_eq!(nested.loop_depth(), 1);
/// ```