aicx 0.10.0

Context search and retrieval for AI-agent session history — find relevant past events, recover user intent, identify unfinished deliverables, and spot claims never backed by results
Documentation
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
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
# Vetcoders Agent Operating Guide v1

<!-- loctree-advise: v0.12 -->

> Loctree gives **sight**.
>
> AICX gives **insight**.
>
> 𝚅𝚒𝚋𝚎𝚌𝚛𝚊𝚏𝚝𝚎𝚍. gives **hands** to craft products.

## Loctree Operating Rule

> For structural orientation start with Loctree.
> (**NEW!** `>=0.12`: Now also for literal occurrences).

This repository should be treated as a structural living system. To help you
to not treat it as a loose pile of files and to boost velocity without
loosing orientation in complexity you are equipped with advanced AST engine
represented by `loctree-mcp` and `loct` cli tool.

Before making structural assumptions, inspect the map.

Before changing behavior, understand impact.

Before creating new symbols, check whether the shape already exists.

Loctree is the default structural map at **session start** and for further
repository work. It makes dependencies, blast radius, symbols, runtime entry
points, dead surfaces, duplicates, and exact occurrences visible faster than
manual rummaging.

### Recommended daily order:

1. Loctree MCP
2. Loct CLI
3. Local tools (`rg`, `grep`, `awk`, `sed`, `cat`)
4. Manual inspection

### Before editing

| Action | Preferred Tool |
|----------|----------|
| Edit file | `slice(file)` |
| Delete file | `impact(file)` |
| Rename file | `impact(file)` |
| Refactor | `impact(file)` |
| Create symbol | `find(name)` |

### Literal truth before assumptions

Use:

- `find --literal`
- `loct occurrences IDENTIFIER`
- `loct body SYMBOL`

before broad text search.

### Fallbacks are allowed

Use:

- `rg`
- `grep`
- `awk`
- `sed`
- `cat`

when Loctree cannot answer cleanly.

### Loctree Feedback Loop

For external users and downstream adopters, `loctree-fail.md` is the intake
channel for bugs, missing language coverage, UX problems, and fallback cases.
They append enough evidence for the Loctree team to reproduce and prioritize.

For Loctree contributors working inside this repository, an observed or reported
Loctree failure is product work, not note-taking. The order is:

1. Reproduce the failure.
2. Add a focused test or contract.
3. Fix it in the same cut when the scope is bounded.
4. Append to `loctree-fail.md` only when repair cannot land now, and state the
   concrete blocker.

If Loctree is:

- wrong (blind, false positive, misleading)
- stale (even after manual force rescan)
- slow (for 200k LOC repositories)
- awkward (proper but noisy or mishaped)
- missing language support (despite claiming support)
- missing an important surface (feature request)
- suggesting an obvious improvement (nitpicks)

check the `--version` you currently use and append a note to:

`.loctree/loctree-fail.md`

> Be aware that version check is crucial, as ome functions like
> `find --literal` are available in `v0.12.x` or upper.

#### Rules

- Never recreate the file.
- Never overwrite the file.
- Always append.

#### Valid entries

- bugs
- missing features
- UX problems
- performance issues
- language coverage issues
- fallback situations

Do not ommit the entry because you sweeped the file and already found
a similar case descriptions. Repeated reports are signal, not noise.

---

# Why This Matters

Loctree changes agent work from:

> text rummaging

into:

> map-first engineering

The goal is not obedience.

The goal is:

- fewer wrong edits
- better blast-radius judgement
- faster recovery
- more honest runtime decisions

<!-- /loctree-advise -->

---

# Agent Behavior Standard

## Work From Structure Before Text

Start with `vc-init`. Do not assume repository shape from filenames alone.

Always identify:

- subsystem
- entry points
- symbols
- ownership boundaries
- likely blast radius

Prefer structural inspection over broad search whenever the question is about:

- dependency
- ownership
- impact
- location

You can use raw text search even when:

- the question is literal
- the question is local

You gain beautiufly curated context around your search. If `loctree-mcp` or
`loct` cli fail, report it honestly and fall back into `rg`, `grep`, `awk`,
`sed` or any tool you are familiar with.

---

## Do Not Edit Blind

Before modifying code:

1. Locate the target.
2. Inspect local implementation.
3. Inspect callers and dependents.
4. Check nearby tests, examples, and docs.
5. Make the smallest coherent change.
6. Verify through the closest runtime path.

If verification cannot be run:

> Say so explicitly.

---

## Do Not Create Parallel Systems Casually

Before introducing:

- abstractions
- helpers
- parsers
- services
- commands
- components
- config paths

check whether one already exists.

If you introduce a new path:

> Explain why reuse was incorrect.

Avoid duplicate systems created only because the agent did not look hard enough.

---

## Prefer Runtime Truth

Static structure matters.

Runtime behavior decides.

When changing:

- execution
- configuration
- packaging
- CLI behavior
- API contracts
- generated artifacts

verify against the real execution path whenever possible.

Passing type checks is useful.

It is not the same thing as product readiness.

---

## Keep The Repository Legible

Prefer changes that improve understanding.

Avoid cleverness that hides shape.

Preserve naming consistency.

Do not bury important behavior inside glue code.

If a file becomes a dumping ground:

> Call it out.

---

## Respect Existing Work

Do not:

- revert
- delete
- rewrite

code you do not understand.

Do not assume unfamiliar changes are safe to discard.

If the repository is moving:

> Re-read before acting.

Treat concurrent agents or human work as part of the system.

---

## Use Direct Language In Handoffs

Always state:

- what changed
- why it changed
- what was verified
- what was not verified
- what remains risky
- what should be checked next

Do not hide uncertainty.

Do not claim confidence you have not earned.

---

## AICX Problem Log

Every AICX problem observed while working in this repository must be appended
to the operator-managed log:

`~/.aicx/aicx-problems.md`

Problem means: bug, regression risk, flaky behavior, contract drift,
docs/runtime mismatch, tooling failure, test gap, zombie path, unsafe fallback,
or a working decision likely to return as product debt.

Rules:

- Append at the end. Never overwrite or reorganize the log.
- Do not skip an entry because a similar one may already exist. Repetition is
  signal.
- Do not store secrets, tokens, PII, PHI, full private payloads, or private
  customer/session material. Redact and describe the shape instead.
- Keep Loctree tool failures in `.loctree/loctree-fail.md`; keep AICX product
  and runtime problems in the AICX problem log above.
- Preferred helper:
  `tools/aicx_problem_log.sh "short title" <<'EOF' ... EOF`
- If the helper fails, append manually. Missing helper is not a reason to skip
  the log.
# Vetcoders Agent Operating Guide v1

<!-- loctree-advise: v0.12 -->

> Loctree gives **sight**.
>
> AICX gives **insight**.
>
> 𝚅𝚒𝚋𝚎𝚌𝚛𝚊𝚏𝚝𝚎𝚍. gives **hands** to craft products.

## Loctree Operating Rule

> For structural orientation start with Loctree.
> (**NEW!** `>=0.12`: Now also for literal occurrences).

This repository should be treated as a structural living system. To help you
to not treat it as a loose pile of files and to boost velocity without
loosing orientation in complexity you are equipped with advanced AST engine
represented by `loctree-mcp` and `loct` cli tool.

Before making structural assumptions, inspect the map.

Before changing behavior, understand impact.

Before creating new symbols, check whether the shape already exists.

Loctree is the default structural map at **session start** and for further
repository work. It makes dependencies, blast radius, symbols, runtime entry
points, dead surfaces, duplicates, and exact occurrences visible faster than
manual rummaging.

### Recommended daily order:

1. Loctree MCP
2. Loct CLI
3. Local tools (`rg`, `grep`, `awk`, `sed`, `cat`)
4. Manual inspection

### Before editing

| Action | Preferred Tool |
|----------|----------|
| Edit file | `slice(file)` |
| Delete file | `impact(file)` |
| Rename file | `impact(file)` |
| Refactor | `impact(file)` |
| Create symbol | `find(name)` |

### Literal truth before assumptions

Use:

- `find --literal`
- `loct occurrences IDENTIFIER`
- `loct body SYMBOL`

before broad text search.

### Fallbacks are allowed

Use:

- `rg`
- `grep`
- `awk`
- `sed`
- `cat`

when Loctree cannot answer cleanly.

### Loctree Feedback Loop

For external users and downstream adopters, `loctree-fail.md` is the intake
channel for bugs, missing language coverage, UX problems, and fallback cases.
They append enough evidence for the Loctree team to reproduce and prioritize.

For Loctree contributors working inside this repository, an observed or reported
Loctree failure is product work, not note-taking. The order is:

1. Reproduce the failure.
2. Add a focused test or contract.
3. Fix it in the same cut when the scope is bounded.
4. Append to `loctree-fail.md` only when repair cannot land now, and state the
   concrete blocker.

If Loctree is:

- wrong (blind, false positive, misleading)
- stale (even after manual force rescan)
- slow (for 200k LOC repositories)
- awkward (proper but noisy or mishaped)
- missing language support (despite claiming support)
- missing an important surface (feature request)
- suggesting an obvious improvement (nitpicks)

check the `--version` you currently use and append a note to:

`.loctree/loctree-fail.md`

> Be aware that version check is crucial, as ome functions like
> `find --literal` are available in `v0.12.x` or upper.

#### Rules

- Never recreate the file.
- Never overwrite the file.
- Always append.

#### Valid entries

- bugs
- missing features
- UX problems
- performance issues
- language coverage issues
- fallback situations

Do not ommit the entry because you sweeped the file and already found
a similar case descriptions. Repeated reports are signal, not noise.

---

# Why This Matters

Loctree changes agent work from:

> text rummaging

into:

> map-first engineering

The goal is not obedience.

The goal is:

- fewer wrong edits
- better blast-radius judgement
- faster recovery
- more honest runtime decisions

<!-- /loctree-advise -->

---

# Agent Behavior Standard

## Work From Structure Before Text

Start with `vc-init`. Do not assume repository shape from filenames alone.

Always identify:

- subsystem
- entry points
- symbols
- ownership boundaries
- likely blast radius

Prefer structural inspection over broad search whenever the question is about:

- dependency
- ownership
- impact
- location

You can use raw text search even when:

- the question is literal
- the question is local

You gain beautiufly curated context around your search. If `loctree-mcp` or
`loct` cli fail, report it honestly and fall back into `rg`, `grep`, `awk`,
`sed` or any tool you are familiar with.

---

## Do Not Edit Blind

Before modifying code:

1. Locate the target.
2. Inspect local implementation.
3. Inspect callers and dependents.
4. Check nearby tests, examples, and docs.
5. Make the smallest coherent change.
6. Verify through the closest runtime path.

If verification cannot be run:

> Say so explicitly.

---

## Do Not Create Parallel Systems Casually

Before introducing:

- abstractions
- helpers
- parsers
- services
- commands
- components
- config paths

check whether one already exists.

If you introduce a new path:

> Explain why reuse was incorrect.

Avoid duplicate systems created only because the agent did not look hard enough.

---

## Prefer Runtime Truth

Static structure matters.

Runtime behavior decides.

When changing:

- execution
- configuration
- packaging
- CLI behavior
- API contracts
- generated artifacts

verify against the real execution path whenever possible.

Passing type checks is useful.

It is not the same thing as product readiness.

---

## Keep The Repository Legible

Prefer changes that improve understanding.

Avoid cleverness that hides shape.

Preserve naming consistency.

Do not bury important behavior inside glue code.

If a file becomes a dumping ground:

> Call it out.

---

## Respect Existing Work

Do not:

- revert
- delete
- rewrite

code you do not understand.

Do not assume unfamiliar changes are safe to discard.

If the repository is moving:

> Re-read before acting.

Treat concurrent agents or human work as part of the system.

---

## Use Direct Language In Handoffs

Always state:

- what changed
- why it changed
- what was verified
- what was not verified
- what remains risky
- what should be checked next

Do not hide uncertainty.

Do not claim confidence you have not earned.