@teevee98/synapse-hub 4.1.1 → 4.1.2
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/.agents/rules/systematic-debugging.md +1 -0
- package/AGENTS.md +2 -1
- package/README.en.md +2 -2
- package/README.md +2 -2
- package/agentskills.json +1 -1
- package/docs/ARCHITECTURE.en.md +1 -1
- package/docs/ARCHITECTURE.md +1 -1
- package/package.json +1 -1
- package/plugin.json +1 -1
- package/skills_manifest.json +1 -1
- package/synapse.config.json +1 -1
- package/tools/.infer_baseline.json +773 -5
- package/tools/__init__.py +1 -1
- package/tools/benchmark_swarm_context.py +4 -1
- package/tools/browser_helpers.js +2 -2
- package/tools/cli.py +3 -3
- package/tools/core/token_estimator.py +120 -5
- package/tools/dashboard_generator.py +6 -6
- package/tools/distill_log.py +73 -22
- package/tools/infer_linter.py +6 -5
- package/tools/mcp_docx_server.py +1 -1
- package/tools/mcp_excel_server.py +1 -1
- package/tools/mcp_server.py +1 -1
- package/tools/rules_router.py +66 -24
- package/tools/ui_server.py +1 -1
- package/tools/update_sentinel.py +3 -3
- package/tools/welcome_page.py +1 -1
package/tools/__init__.py
CHANGED
|
@@ -414,7 +414,10 @@ def format_markdown_report(bench: Dict[str, Any]) -> str:
|
|
|
414
414
|
"",
|
|
415
415
|
"## Track 2: Empirical Reference Suites Benchmark (Real Projects)",
|
|
416
416
|
"",
|
|
417
|
-
"
|
|
417
|
+
"> *Methodology: Dynamically enumerates all 13 rules and 53 skills in repository without static per-skill multiplier assumptions.*",
|
|
418
|
+
"> *Disclosure: 'Avoided Disk Read' measures disk file bytes not read from storage, distinguished from prompt context tokens.*",
|
|
419
|
+
"",
|
|
420
|
+
"| Reference Project | Detected Stack | Confidence | Catalog Summary Saved | Full Hydration Saved | Avoided Disk Read (I/O) |",
|
|
418
421
|
"|---|---|:---:|:---:|:---:|:---:|",
|
|
419
422
|
])
|
|
420
423
|
for p in emp["results"]:
|
package/tools/browser_helpers.js
CHANGED
|
@@ -13,7 +13,7 @@
|
|
|
13
13
|
(function () {
|
|
14
14
|
'use strict';
|
|
15
15
|
|
|
16
|
-
if (window.__qa && window.__qa.__version === '4.1.
|
|
16
|
+
if (window.__qa && window.__qa.__version === '4.1.2') {
|
|
17
17
|
return window.__qa;
|
|
18
18
|
}
|
|
19
19
|
|
|
@@ -47,7 +47,7 @@
|
|
|
47
47
|
}
|
|
48
48
|
|
|
49
49
|
const qa = {
|
|
50
|
-
__version: '4.1.
|
|
50
|
+
__version: '4.1.2',
|
|
51
51
|
|
|
52
52
|
/**
|
|
53
53
|
* 1. Smart Action Batching
|
package/tools/cli.py
CHANGED
|
@@ -99,7 +99,7 @@ def cmd_set_lang(args):
|
|
|
99
99
|
def cmd_version(args):
|
|
100
100
|
"""Displays the current hub version, skill catalog count, and license information."""
|
|
101
101
|
manifest_path = os.path.join(REPO_ROOT, "skills_manifest.json")
|
|
102
|
-
version = "4.1.
|
|
102
|
+
version = "4.1.2"
|
|
103
103
|
if os.path.exists(manifest_path):
|
|
104
104
|
try:
|
|
105
105
|
import json
|
|
@@ -686,14 +686,14 @@ def cmd_trend(args):
|
|
|
686
686
|
|
|
687
687
|
def _hub_version():
|
|
688
688
|
manifest_path = os.path.join(REPO_ROOT, "skills_manifest.json")
|
|
689
|
-
version = "4.1.
|
|
689
|
+
version = "4.1.2"
|
|
690
690
|
if os.path.exists(manifest_path):
|
|
691
691
|
try:
|
|
692
692
|
import json
|
|
693
693
|
with open(manifest_path, "r", encoding="utf-8") as f:
|
|
694
694
|
version = json.load(f).get("version", version)
|
|
695
695
|
except (OSError, json.JSONDecodeError):
|
|
696
|
-
version = "4.1.
|
|
696
|
+
version = "4.1.2"
|
|
697
697
|
return version
|
|
698
698
|
|
|
699
699
|
|
|
@@ -16,10 +16,13 @@ Key Capabilities:
|
|
|
16
16
|
- Non-ASCII / Vietnamese diacritics / Unicode: ~1.8 characters/token.
|
|
17
17
|
2. Optional native BPE tokenization if tiktoken is available in runtime.
|
|
18
18
|
3. Structured token breakdown (Rules, Catalog Metadata, Full Skill Bodies, System Prefix).
|
|
19
|
+
4. Real dynamic baseline calculation enumerating 100% of skills & rules without static estimates.
|
|
19
20
|
"""
|
|
20
21
|
|
|
21
22
|
from __future__ import annotations
|
|
22
23
|
|
|
24
|
+
import glob
|
|
25
|
+
import json
|
|
23
26
|
import os
|
|
24
27
|
import re
|
|
25
28
|
from typing import Any, Dict, List, Optional
|
|
@@ -36,6 +39,21 @@ try:
|
|
|
36
39
|
except ImportError:
|
|
37
40
|
_TIKTOKEN_ENCODER = None
|
|
38
41
|
|
|
42
|
+
_SKILLS_MANIFEST_CACHE: Optional[Dict[str, Dict[str, Any]]] = None
|
|
43
|
+
|
|
44
|
+
|
|
45
|
+
def get_estimator_info() -> Dict[str, Any]:
|
|
46
|
+
"""Returns active tokenizer engine metadata, exactness status, and model hint."""
|
|
47
|
+
return {
|
|
48
|
+
"engine": "cl100k_base" if _TIKTOKEN_ENCODER is not None else "heuristic_polyglot",
|
|
49
|
+
"is_exact": _TIKTOKEN_ENCODER is not None,
|
|
50
|
+
"encoding": "cl100k_base",
|
|
51
|
+
"calibrated_ratios": {
|
|
52
|
+
"ascii_chars_per_token": 3.85,
|
|
53
|
+
"non_ascii_chars_per_token": 1.80,
|
|
54
|
+
},
|
|
55
|
+
}
|
|
56
|
+
|
|
39
57
|
|
|
40
58
|
def count_tokens(text: str, model_hint: str = "cl100k") -> int:
|
|
41
59
|
"""
|
|
@@ -81,6 +99,38 @@ def count_file_tokens(file_path: str) -> int:
|
|
|
81
99
|
return 0
|
|
82
100
|
|
|
83
101
|
|
|
102
|
+
def _get_skills_manifest(repo_root: str) -> Dict[str, Dict[str, Any]]:
|
|
103
|
+
"""Loads and caches skills_manifest.json mapping skill_id to metadata."""
|
|
104
|
+
global _SKILLS_MANIFEST_CACHE
|
|
105
|
+
if _SKILLS_MANIFEST_CACHE is not None:
|
|
106
|
+
return _SKILLS_MANIFEST_CACHE
|
|
107
|
+
|
|
108
|
+
manifest_path = os.path.join(repo_root, "skills_manifest.json")
|
|
109
|
+
_SKILLS_MANIFEST_CACHE = {}
|
|
110
|
+
if os.path.isfile(manifest_path):
|
|
111
|
+
try:
|
|
112
|
+
with open(manifest_path, "r", encoding="utf-8") as f:
|
|
113
|
+
data = json.load(f)
|
|
114
|
+
for s in data.get("skills", []):
|
|
115
|
+
sid = s.get("id")
|
|
116
|
+
if sid:
|
|
117
|
+
_SKILLS_MANIFEST_CACHE[sid] = s
|
|
118
|
+
except (OSError, json.JSONDecodeError) as err:
|
|
119
|
+
_ = err
|
|
120
|
+
return _SKILLS_MANIFEST_CACHE
|
|
121
|
+
|
|
122
|
+
|
|
123
|
+
def serialize_skill_summary(skill_name: str, meta: Optional[Dict[str, Any]] = None) -> str:
|
|
124
|
+
"""Formats a concise catalog summary line for a skill."""
|
|
125
|
+
if meta:
|
|
126
|
+
name = meta.get("name", skill_name)
|
|
127
|
+
domain = meta.get("domain", "")
|
|
128
|
+
desc = meta.get("description") or meta.get("summary", "")
|
|
129
|
+
domain_str = f" [{domain}]" if domain else ""
|
|
130
|
+
return f"- {name} ({skill_name}){domain_str}: {desc}"
|
|
131
|
+
return f"- {skill_name}: Enterprise QA Automation Skill"
|
|
132
|
+
|
|
133
|
+
|
|
84
134
|
def calculate_hydration_payload(
|
|
85
135
|
repo_root: str,
|
|
86
136
|
rules: List[str],
|
|
@@ -88,11 +138,12 @@ def calculate_hydration_payload(
|
|
|
88
138
|
) -> Dict[str, Any]:
|
|
89
139
|
"""
|
|
90
140
|
Computes precise dual-tier token metrics for a set of selected rules and skills:
|
|
91
|
-
1. summary: Catalog metadata only (name +
|
|
141
|
+
1. summary: Catalog metadata only (serialized name + domain + description).
|
|
92
142
|
2. full: Complete SKILL.md body hydrated into context.
|
|
93
143
|
"""
|
|
94
144
|
rules_dir = os.path.join(repo_root, ".agents", "rules")
|
|
95
145
|
skills_root = os.path.join(repo_root, ".agents", "skills")
|
|
146
|
+
manifest = _get_skills_manifest(repo_root)
|
|
96
147
|
|
|
97
148
|
rule_tokens = 0
|
|
98
149
|
rule_bytes = 0
|
|
@@ -105,12 +156,15 @@ def calculate_hydration_payload(
|
|
|
105
156
|
rule_tokens += 500 # Default estimate for virtual rule
|
|
106
157
|
rule_bytes += 2000
|
|
107
158
|
|
|
108
|
-
|
|
159
|
+
catalog_lines = []
|
|
109
160
|
full_skill_tokens = 0
|
|
110
161
|
full_skill_bytes = 0
|
|
111
162
|
|
|
112
163
|
# Locate each skill directory in .agents/skills
|
|
113
164
|
for skill_name in skills:
|
|
165
|
+
meta = manifest.get(skill_name)
|
|
166
|
+
catalog_lines.append(serialize_skill_summary(skill_name, meta))
|
|
167
|
+
|
|
114
168
|
skill_file = _find_skill_file(skills_root, skill_name)
|
|
115
169
|
if skill_file and os.path.isfile(skill_file):
|
|
116
170
|
tokens = count_file_tokens(skill_file)
|
|
@@ -118,9 +172,11 @@ def calculate_hydration_payload(
|
|
|
118
172
|
full_skill_tokens += tokens
|
|
119
173
|
full_skill_bytes += size
|
|
120
174
|
else:
|
|
121
|
-
# Fallback average skill size
|
|
122
|
-
full_skill_tokens +=
|
|
123
|
-
full_skill_bytes +=
|
|
175
|
+
# Fallback average skill size if file missing
|
|
176
|
+
full_skill_tokens += 1200
|
|
177
|
+
full_skill_bytes += 4500
|
|
178
|
+
|
|
179
|
+
catalog_skill_tokens = count_tokens("\n".join(catalog_lines)) if catalog_lines else 0
|
|
124
180
|
|
|
125
181
|
return {
|
|
126
182
|
"rule_tokens": rule_tokens,
|
|
@@ -134,6 +190,65 @@ def calculate_hydration_payload(
|
|
|
134
190
|
}
|
|
135
191
|
|
|
136
192
|
|
|
193
|
+
def get_full_unrouted_baseline(repo_root: str) -> Dict[str, Any]:
|
|
194
|
+
"""
|
|
195
|
+
Enumerates and calculates actual token and byte counts for 100% of all rules in .agents/rules
|
|
196
|
+
and 100% of all skills in .agents/skills.
|
|
197
|
+
Guarantees zero static assumptions (replaces 53 * 1800 and 53 * 120 with exact file token counts).
|
|
198
|
+
"""
|
|
199
|
+
rules_dir = os.path.join(repo_root, ".agents", "rules")
|
|
200
|
+
skills_root = os.path.join(repo_root, ".agents", "skills")
|
|
201
|
+
manifest = _get_skills_manifest(repo_root)
|
|
202
|
+
|
|
203
|
+
# 1. Enumerate all rule files
|
|
204
|
+
all_rule_files: List[str] = []
|
|
205
|
+
if os.path.isdir(rules_dir):
|
|
206
|
+
all_rule_files = sorted([f for f in os.listdir(rules_dir) if f.endswith(".md")])
|
|
207
|
+
|
|
208
|
+
total_rule_tokens = 0
|
|
209
|
+
total_rule_bytes = 0
|
|
210
|
+
for rf in all_rule_files:
|
|
211
|
+
p = os.path.join(rules_dir, rf)
|
|
212
|
+
total_rule_tokens += count_file_tokens(p)
|
|
213
|
+
total_rule_bytes += os.path.getsize(p)
|
|
214
|
+
|
|
215
|
+
# 2. Enumerate all 53 SKILL.md files
|
|
216
|
+
skill_files = sorted(glob.glob(os.path.join(skills_root, "*", "*", "SKILL.md")))
|
|
217
|
+
total_skill_tokens = 0
|
|
218
|
+
total_skill_bytes = 0
|
|
219
|
+
all_skill_ids: List[str] = []
|
|
220
|
+
|
|
221
|
+
for sf in skill_files:
|
|
222
|
+
total_skill_tokens += count_file_tokens(sf)
|
|
223
|
+
total_skill_bytes += os.path.getsize(sf)
|
|
224
|
+
skill_id = os.path.basename(os.path.dirname(sf))
|
|
225
|
+
all_skill_ids.append(skill_id)
|
|
226
|
+
|
|
227
|
+
# 3. Dynamic Catalog Summary tokens across all skills
|
|
228
|
+
catalog_lines = []
|
|
229
|
+
for sid in all_skill_ids:
|
|
230
|
+
meta = manifest.get(sid)
|
|
231
|
+
catalog_lines.append(serialize_skill_summary(sid, meta))
|
|
232
|
+
|
|
233
|
+
total_cat_tokens = count_tokens("\n".join(catalog_lines)) if catalog_lines else 0
|
|
234
|
+
|
|
235
|
+
return {
|
|
236
|
+
"rule_count": len(all_rule_files),
|
|
237
|
+
"rule_tokens": total_rule_tokens,
|
|
238
|
+
"rule_bytes": total_rule_bytes,
|
|
239
|
+
"skill_count": len(skill_files),
|
|
240
|
+
"skill_full_body_tokens": total_skill_tokens,
|
|
241
|
+
"skill_full_bytes": total_skill_bytes,
|
|
242
|
+
"skill_catalog_tokens": total_cat_tokens,
|
|
243
|
+
"total_summary_baseline": total_rule_tokens + total_cat_tokens,
|
|
244
|
+
"total_full_baseline": total_rule_tokens + total_skill_tokens,
|
|
245
|
+
"total_full_bytes": total_rule_bytes + total_skill_bytes,
|
|
246
|
+
"all_rules": all_rule_files,
|
|
247
|
+
"all_skills": all_skill_ids,
|
|
248
|
+
"estimator_info": get_estimator_info(),
|
|
249
|
+
}
|
|
250
|
+
|
|
251
|
+
|
|
137
252
|
def _find_skill_file(skills_root: str, skill_name: str) -> Optional[str]:
|
|
138
253
|
"""Finds SKILL.md for a given skill name across tiered category folders."""
|
|
139
254
|
if not os.path.isdir(skills_root):
|
|
@@ -1348,7 +1348,7 @@ def _build_target_translations(
|
|
|
1348
1348
|
"th_step_status": "Trạng Thái",
|
|
1349
1349
|
"th_step_details": "Chi Tiết Kỹ Thuật (OTel & AST)",
|
|
1350
1350
|
|
|
1351
|
-
"footer_text": "Báo cáo được khởi tạo tự động bởi Synapse-Hub v4.1.
|
|
1351
|
+
"footer_text": "Báo cáo được khởi tạo tự động bởi Synapse-Hub v4.1.2 • Bản quyền © Thế Vị (@TeeVee98) • Chuẩn Clean POM & Kiểm Thử Tự Chủ",
|
|
1352
1352
|
}
|
|
1353
1353
|
|
|
1354
1354
|
translations_en = {
|
|
@@ -1426,7 +1426,7 @@ def _build_target_translations(
|
|
|
1426
1426
|
"th_step_status": "Status",
|
|
1427
1427
|
"th_step_details": "Technical Details (OTel & AST)",
|
|
1428
1428
|
|
|
1429
|
-
"footer_text": "Generated automatically by Synapse-Hub v4.1.
|
|
1429
|
+
"footer_text": "Generated automatically by Synapse-Hub v4.1.2 • Copyright © Thế Vị (@TeeVee98) • Clean POM & Autonomous Testing",
|
|
1430
1430
|
}
|
|
1431
1431
|
|
|
1432
1432
|
_inject_gate_translations(translations_vi, translations_en, gates)
|
|
@@ -1915,7 +1915,7 @@ def _build_hub_translations(
|
|
|
1915
1915
|
"""Builds bilingual dictionaries for Hub Dashboard."""
|
|
1916
1916
|
translations_vi = {
|
|
1917
1917
|
"header_title": "Bảng Điều Hành Chất Lượng Doanh Nghiệp",
|
|
1918
|
-
"header_subtitle": "Synapse-Hub v4.1.
|
|
1918
|
+
"header_subtitle": "Synapse-Hub v4.1.2 • Hệ Sinh Thái Kiểm Thử AI Agent Tự Chủ",
|
|
1919
1919
|
"nav_welcome": "Giới Thiệu",
|
|
1920
1920
|
"meta_mode": "🏷️ Chế độ: <strong>HỆ SINH THÁI QA HUB</strong>",
|
|
1921
1921
|
"meta_gen": f"🕒 Khởi tạo: <strong>{now_str}</strong>",
|
|
@@ -1985,12 +1985,12 @@ def _build_hub_translations(
|
|
|
1985
1985
|
"th_sk_ver": "Phiên Bản",
|
|
1986
1986
|
"th_sk_cap": "Khả Năng Nghiệp Vụ Cốt Lõi",
|
|
1987
1987
|
|
|
1988
|
-
"footer_text": "Bảng Điều Hành Chất Lượng Doanh Nghiệp • Synapse-Hub v4.1.
|
|
1988
|
+
"footer_text": "Bảng Điều Hành Chất Lượng Doanh Nghiệp • Synapse-Hub v4.1.2 • Bản quyền © Thế Vị (@TeeVee98) • Hệ Thống Kiểm Thử AI Agent Tự Chủ",
|
|
1989
1989
|
}
|
|
1990
1990
|
|
|
1991
1991
|
translations_en = {
|
|
1992
1992
|
"header_title": "Enterprise QA Executive Dashboard",
|
|
1993
|
-
"header_subtitle": "Synapse-Hub v4.1.
|
|
1993
|
+
"header_subtitle": "Synapse-Hub v4.1.2 • Autonomous AI Agent Testing Ecosystem",
|
|
1994
1994
|
"nav_welcome": "Welcome",
|
|
1995
1995
|
"meta_mode": "🏷️ Mode: <strong>QA HUB ECOSYSTEM</strong>",
|
|
1996
1996
|
"meta_gen": f"🕒 Generated: <strong>{now_str}</strong>",
|
|
@@ -2060,7 +2060,7 @@ def _build_hub_translations(
|
|
|
2060
2060
|
"th_sk_ver": "Version",
|
|
2061
2061
|
"th_sk_cap": "Core Capabilities",
|
|
2062
2062
|
|
|
2063
|
-
"footer_text": "Enterprise QA Executive Dashboard • Synapse-Hub v4.1.
|
|
2063
|
+
"footer_text": "Enterprise QA Executive Dashboard • Synapse-Hub v4.1.2 • Copyright © Thế Vị (@TeeVee98) • Autonomous AI Agent Testing Ecosystem",
|
|
2064
2064
|
}
|
|
2065
2065
|
|
|
2066
2066
|
_inject_gate_translations(translations_vi, translations_en, gates)
|
package/tools/distill_log.py
CHANGED
|
@@ -5,9 +5,18 @@
|
|
|
5
5
|
|
|
6
6
|
"""
|
|
7
7
|
tools/distill_log.py — Enterprise Multi-Framework Log Distiller (Token & Failure Sentinel)
|
|
8
|
-
|
|
9
|
-
|
|
10
|
-
|
|
8
|
+
------------------------------------------------------------------------------------------
|
|
9
|
+
Distills massive, repetitive execution logs (Katalon Studio, Gradle, Playwright, Jest, Pytest)
|
|
10
|
+
down to exact failure triage blocks without false passes, false failures on info logs, or dropped assertion diffs.
|
|
11
|
+
|
|
12
|
+
Key Guarantees:
|
|
13
|
+
1. Case-insensitive failure keyword scanning.
|
|
14
|
+
2. Exit code Truth Anchor (nonzero -> RUNNER_NONZERO_EXIT; missing -> NO_FAILURE_SIGNATURE_DETECTED;
|
|
15
|
+
clean 0 with failure text -> CONFLICTING_SIGNALS).
|
|
16
|
+
3. Informational logging filter ([INFO], [DEBUG]) preventing false-positive assertion detection.
|
|
17
|
+
4. Genuine token_budget enforcement via calibrated tokenizer (default: 800 tokens).
|
|
18
|
+
5. 15-line block bounding and noise suppression, compressing 95% of log volume.
|
|
19
|
+
6. Provenance retention: points back to raw log file for forensic tracing.
|
|
11
20
|
"""
|
|
12
21
|
|
|
13
22
|
from __future__ import annotations
|
|
@@ -26,6 +35,20 @@ if hasattr(sys.stdout, "reconfigure"):
|
|
|
26
35
|
except Exception:
|
|
27
36
|
pass
|
|
28
37
|
|
|
38
|
+
# Ensure repository root is on sys.path
|
|
39
|
+
_REPO_ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
|
|
40
|
+
if _REPO_ROOT not in sys.path:
|
|
41
|
+
sys.path.insert(0, _REPO_ROOT)
|
|
42
|
+
|
|
43
|
+
try:
|
|
44
|
+
from tools.core.token_estimator import count_tokens
|
|
45
|
+
except ImportError:
|
|
46
|
+
try:
|
|
47
|
+
from core.token_estimator import count_tokens
|
|
48
|
+
except ImportError:
|
|
49
|
+
def count_tokens(text: str, model_hint: str = "cl100k") -> int:
|
|
50
|
+
return max(1, len(text) // 4)
|
|
51
|
+
|
|
29
52
|
# Universal noise patterns (passing steps, zero failures summary, timestamps, system initialization)
|
|
30
53
|
NOISE_PATTERNS = [
|
|
31
54
|
re.compile(r"^\s*$"),
|
|
@@ -49,6 +72,12 @@ NOISE_PATTERNS = [
|
|
|
49
72
|
re.compile(r"Katalon Studio.*licensed to"),
|
|
50
73
|
]
|
|
51
74
|
|
|
75
|
+
# Informational logging prefix pattern (prevents informational lines from initiating failure blocks)
|
|
76
|
+
INFO_LOG_LINE = re.compile(
|
|
77
|
+
r"^\s*(?:\[?(?:INFO|DEBUG|NOTICE|TRACE|VERBOSE|LOG)\]?|\d{4}-\d{2}-\d{2}[T\s]\S+\s+\[?(?:INFO|DEBUG|NOTICE|TRACE|VERBOSE|LOG)\]?)\b",
|
|
78
|
+
re.IGNORECASE,
|
|
79
|
+
)
|
|
80
|
+
|
|
52
81
|
# Explicit false-indicator filter (phrases like "without failure", "0 failures")
|
|
53
82
|
PASSING_PHRASES = re.compile(
|
|
54
83
|
r"\b(?:without|zero|0|no)\s+(?:obvious\s+)?(?:failures?|errors?|exceptions?)\b",
|
|
@@ -68,7 +97,7 @@ FAILURE_INDICATORS = [
|
|
|
68
97
|
re.compile(r"FAIL\s+.*(?:\.spec|\.test)", re.IGNORECASE),
|
|
69
98
|
re.compile(r"===+\s*FAILURES\s*===+", re.IGNORECASE),
|
|
70
99
|
re.compile(r"E\s+assert\b", re.IGNORECASE),
|
|
71
|
-
re.compile(r"assert\s+.*==\s*", re.IGNORECASE),
|
|
100
|
+
re.compile(r"(?:^|\s)(?:E\s+|>\s+|AssertionError:\s*)assert\s+.*==\s*", re.IGNORECASE),
|
|
72
101
|
re.compile(r"expect\(.*\)\.", re.IGNORECASE),
|
|
73
102
|
re.compile(r"Timeout \d+ms exceeded", re.IGNORECASE),
|
|
74
103
|
re.compile(r"Traceback \(most recent call last\):", re.IGNORECASE),
|
|
@@ -97,7 +126,7 @@ def distill(
|
|
|
97
126
|
) -> str:
|
|
98
127
|
"""
|
|
99
128
|
Distills raw execution logs down to high-signal failure blocks while guaranteeing
|
|
100
|
-
zero false passes
|
|
129
|
+
zero false passes, zero false-positive assertions on info logs, and enforcing token budgets.
|
|
101
130
|
"""
|
|
102
131
|
raw_bytes = len(log_text.encode("utf-8"))
|
|
103
132
|
lines = log_text.splitlines()
|
|
@@ -121,11 +150,14 @@ def distill(
|
|
|
121
150
|
if m_tc:
|
|
122
151
|
last_seen_tc = m_tc.group(1)
|
|
123
152
|
|
|
124
|
-
#
|
|
125
|
-
|
|
153
|
+
# Informational log lines cannot initiate failure blocks
|
|
154
|
+
is_info = bool(INFO_LOG_LINE.search(line))
|
|
155
|
+
|
|
156
|
+
# Case-insensitive failure check
|
|
157
|
+
is_fail_line = (not is_info or in_failure_block) and any(p.search(line) for p in FAILURE_INDICATORS)
|
|
126
158
|
is_noise = any(p.search(line) for p in NOISE_PATTERNS) and not is_fail_line
|
|
127
159
|
|
|
128
|
-
if is_fail_line:
|
|
160
|
+
if is_fail_line and not is_info:
|
|
129
161
|
if not in_failure_block:
|
|
130
162
|
in_failure_block = True
|
|
131
163
|
current_block = []
|
|
@@ -207,18 +239,24 @@ def distill(
|
|
|
207
239
|
output.append(f"[RAW LOG ARTIFACT] file://{os.path.abspath(raw_log_path)}")
|
|
208
240
|
return "\n".join(output)
|
|
209
241
|
|
|
210
|
-
#
|
|
211
|
-
if exit_code
|
|
242
|
+
# Invariant: Conflicting Signals Handling when exit_code == 0 but failure keywords present
|
|
243
|
+
if exit_code == 0:
|
|
244
|
+
output.append("⚠️ CONFLICTING_SIGNALS [Exit Code: 0]: Test runner exited with code 0, but failure indicators were detected in log.")
|
|
245
|
+
output.append(" (Check whether failures were swallowed by an outer test harness, or if informational text matched failure patterns).")
|
|
246
|
+
elif exit_code is not None:
|
|
212
247
|
exit_badge = f" [Exit Code: {exit_code}]"
|
|
248
|
+
if failing_tcs:
|
|
249
|
+
output.append(f"❌ FAILURES DETECTED{exit_badge}: {len(failing_tcs)} Test Case(s) marked FAILED.")
|
|
250
|
+
else:
|
|
251
|
+
output.append(f"❌ FAILURES / ERRORS DETECTED{exit_badge}: {len(captured_failures)} failure block(s) isolated.")
|
|
213
252
|
else:
|
|
214
|
-
|
|
253
|
+
if failing_tcs:
|
|
254
|
+
output.append(f"❌ FAILURES DETECTED: {len(failing_tcs)} Test Case(s) marked FAILED.")
|
|
255
|
+
else:
|
|
256
|
+
output.append(f"❌ FAILURES / ERRORS DETECTED: {len(captured_failures)} failure block(s) isolated.")
|
|
215
257
|
|
|
216
258
|
if failing_tcs:
|
|
217
|
-
output.append(f"❌ FAILURES DETECTED{exit_badge}: {len(failing_tcs)} Test Case(s) marked FAILED.")
|
|
218
259
|
output.append(f" Targeted TCs: {', '.join(sorted(failing_tcs))}")
|
|
219
|
-
else:
|
|
220
|
-
output.append(f"❌ FAILURES / ERRORS DETECTED{exit_badge}: {len(captured_failures)} failure block(s) isolated.")
|
|
221
|
-
|
|
222
260
|
if failing_locs:
|
|
223
261
|
unique_locs = sorted(set(failing_locs))
|
|
224
262
|
output.append(f" Failing Script Location(s): {', '.join(unique_locs[:5])}")
|
|
@@ -237,7 +275,6 @@ def distill(
|
|
|
237
275
|
block_text = "\n".join(block)
|
|
238
276
|
block_lines = len(block)
|
|
239
277
|
if current_lines + block_lines > max_payload_lines:
|
|
240
|
-
# Take what fits
|
|
241
278
|
allowed = max_payload_lines - current_lines
|
|
242
279
|
if allowed > 2:
|
|
243
280
|
assembled_blocks.append("\n".join(block[:allowed]))
|
|
@@ -249,14 +286,26 @@ def distill(
|
|
|
249
286
|
|
|
250
287
|
output.append("\n---\n".join(assembled_blocks))
|
|
251
288
|
|
|
252
|
-
#
|
|
253
|
-
|
|
254
|
-
|
|
255
|
-
|
|
289
|
+
# Genuine Token Budget Constraint Enforcement
|
|
290
|
+
if token_budget > 0:
|
|
291
|
+
current_tokens = count_tokens("\n".join(output))
|
|
292
|
+
if current_tokens > token_budget:
|
|
293
|
+
truncated = True
|
|
294
|
+
# Pop lines from the end of the output body until within token budget
|
|
295
|
+
while len(output) > 4 and count_tokens("\n".join(output)) > max(20, token_budget - 25):
|
|
296
|
+
output.pop()
|
|
297
|
+
output.append(f"... [truncated: constrained by token_budget={token_budget} tokens] ...")
|
|
298
|
+
|
|
299
|
+
# Measure distilled body before appending metadata footer
|
|
300
|
+
body_lines = len(output)
|
|
301
|
+
distilled_body_text = "\n".join(output)
|
|
302
|
+
distilled_body_bytes = len(distilled_body_text.encode("utf-8"))
|
|
303
|
+
distilled_tokens = count_tokens(distilled_body_text)
|
|
304
|
+
reduction_pct = round(((raw_bytes - distilled_body_bytes) / max(1, raw_bytes)) * 100, 1)
|
|
256
305
|
|
|
257
306
|
output.append(
|
|
258
|
-
f"\n[DISTILLER METRICS] Distilled: {
|
|
259
|
-
f"Reduction: {reduction_pct}% | Truncated: {truncated}"
|
|
307
|
+
f"\n[DISTILLER METRICS] Distilled Body: {body_lines} lines ({distilled_body_bytes:,} bytes, ~{distilled_tokens} tokens) | "
|
|
308
|
+
f"Reduction: {reduction_pct}% | Truncated: {truncated} | Token Budget: {token_budget}"
|
|
260
309
|
)
|
|
261
310
|
|
|
262
311
|
if raw_log_path:
|
|
@@ -324,6 +373,7 @@ def main():
|
|
|
324
373
|
parser.add_argument("--exit-code", "-e", type=int, default=None, help="Process exit code from test runner.")
|
|
325
374
|
parser.add_argument("--save-raw", "-s", default=None, help="Path to preserve raw log.")
|
|
326
375
|
parser.add_argument("--max-lines", "-m", type=int, default=40, help="Maximum number of output lines.")
|
|
376
|
+
parser.add_argument("--token-budget", "-b", type=int, default=800, help="Maximum token budget for distilled output.")
|
|
327
377
|
parser.add_argument("--shrink", action="store_true", help="Apply caveman/shrink token compression.")
|
|
328
378
|
args = parser.parse_args()
|
|
329
379
|
|
|
@@ -353,6 +403,7 @@ def main():
|
|
|
353
403
|
max_lines=args.max_lines,
|
|
354
404
|
exit_code=args.exit_code,
|
|
355
405
|
raw_log_path=args.save_raw,
|
|
406
|
+
token_budget=args.token_budget,
|
|
356
407
|
)
|
|
357
408
|
if args.shrink:
|
|
358
409
|
summary = shrink_tokens(summary)
|
package/tools/infer_linter.py
CHANGED
|
@@ -86,11 +86,12 @@ ASSERTION_PATTERN = re.compile(
|
|
|
86
86
|
)
|
|
87
87
|
|
|
88
88
|
# Rule L1 & L5: Empty catch / swallowed exceptions across JS/TS/Groovy/Java and Python
|
|
89
|
-
# Also catches try/catch that only console.log/logger.error
|
|
89
|
+
# Also catches try/catch that only console.log/logger.error, promise .catch(() => {}),
|
|
90
|
+
# or dummy assignment bypasses like `_ = err`, `_ = e`, `_ = False`, `const _ = err;`
|
|
90
91
|
EMPTY_CATCH_PATTERN = re.compile(
|
|
91
|
-
r'(?:catch\s*(?:\([^\)]*\))?\s*\{\s*(?:(?:console|logger|log)\.(?:log|warn|error|info)\([^)]*\);?\s*)?\}'
|
|
92
|
-
r'|\.catch\s*\(\s*(?:\([^)]*\)|[a-zA-Z0-9_]+)?\s*=>\s*\{\s*(?:(?:console|logger|log)\.(?:log|warn|error|info)\([^)]*\);?\s*)?\}\s*\)'
|
|
93
|
-
r'|\bexcept(?:\s+[^:]*)?:\s*(?:pass
|
|
92
|
+
r'(?:catch\s*(?:\([^\)]*\))?\s*\{\s*(?:(?:console|logger|log)\.(?:log|warn|error|info)\([^)]*\);?\s*|(?:const|let|var|def)?\s*_\s*=\s*[^;]+;?\s*)?\}'
|
|
93
|
+
r'|\.catch\s*\(\s*(?:\([^)]*\)|[a-zA-Z0-9_]+)?\s*=>\s*\{\s*(?:(?:console|logger|log)\.(?:log|warn|error|info)\([^)]*\);?\s*|(?:const|let|var|def)?\s*_\s*=\s*[^;]+;?\s*)?\}\s*\)'
|
|
94
|
+
r'|\bexcept(?:\s+[^:]*)?:\s*(?:pass|\.\.\.|_\s*=\s*[^#\n;]+)(?:\s*$|\s*#))',
|
|
94
95
|
re.MULTILINE
|
|
95
96
|
)
|
|
96
97
|
|
|
@@ -426,7 +427,7 @@ def analyze_script(file_path, baseline_set=None, disabled_rules=None):
|
|
|
426
427
|
"type": "SWALLOWED_EXCEPTION",
|
|
427
428
|
"severity": "CRITICAL",
|
|
428
429
|
"line": line_no,
|
|
429
|
-
"message": "Empty catch block detected. Swallowed exceptions cause false-pass test runs."
|
|
430
|
+
"message": "Empty catch block or dummy suppression (_ = err) detected. Swallowed exceptions cause false-pass test runs."
|
|
430
431
|
})
|
|
431
432
|
|
|
432
433
|
# Rule 3: Flaky Static Delays (Playwright, Cypress, Pytest, Katalon, Java/Groovy)
|
package/tools/mcp_docx_server.py
CHANGED
|
@@ -60,7 +60,7 @@ if hasattr(sys.stdout, 'reconfigure'):
|
|
|
60
60
|
except Exception:
|
|
61
61
|
pass
|
|
62
62
|
|
|
63
|
-
SERVER_VERSION = "4.1.
|
|
63
|
+
SERVER_VERSION = "4.1.2"
|
|
64
64
|
MAX_DOCX_SIZE = 50 * 1024 * 1024 # 50 MB
|
|
65
65
|
MAX_UNCOMPRESSED_SIZE = 200 * 1024 * 1024 # 200 MB
|
|
66
66
|
MAX_RATIO = 100 # 100:1 compression ratio limit
|
|
@@ -52,7 +52,7 @@ if hasattr(sys.stdout, 'reconfigure'):
|
|
|
52
52
|
except Exception:
|
|
53
53
|
pass
|
|
54
54
|
|
|
55
|
-
SERVER_VERSION = "4.1.
|
|
55
|
+
SERVER_VERSION = "4.1.2"
|
|
56
56
|
MAX_XLSX_SIZE = 50 * 1024 * 1024 # 50 MB
|
|
57
57
|
MAX_UNCOMPRESSED_SIZE = 200 * 1024 * 1024 # 200 MB
|
|
58
58
|
MAX_RATIO = 100 # 100:1 compression ratio limit
|