iframe-proxy | Sunbelt Computer Software

Repository navigation

History

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

#!/usr/bin/env python3

"""

End-to-end self-improvement loop test.

Validates the full CortexOS cycle:

profile → context → digest → score → focus → learn → improve

+ deduplication, tag inference, spaced repetition

"""

import json

import sys

from pathlib import Path

sys.path.insert(0, str(Path(__file__).resolve().parent.parent))

from cortex_core.engine import CortexEngine

PASS = 0

FAIL = 0

def check(label: str, condition: bool, detail: str = ""):

global PASS, FAIL

icon = "✓" if condition else "✗"

status = "PASS" if condition else "FAIL"

if condition:

PASS += 1

else:

FAIL += 1

msg = f" {icon} [{status}] {label}"

if detail:

msg += f" ({detail})"

print(msg)

def section(title: str):

print(f"\n{'=' * 60}")

print(f" {title}")

print(f"{'=' * 60}")

def main():

engine = CortexEngine()

# ── 1. Profile: set user context ────────────────────────

section("1. SET USER PROFILE")

engine.update_profile(

name="Pierre-Henry Soria",

goals=[

"Ship CortexOS as flagship project",

"Build AI-native productivity tools",

"Become visible as AI systems engineer",

interests=[

"AI systems",

"context engines",

"retrieval augmented generation",

"developer productivity",

"healthy living",

"Swift",

"Python",

current_projects=["CortexOS", "pH7Programming YouTube", "pierrewriter.com"],

constraints=["Solo founder", "Sydney timezone"],

ignored_topics=["crypto", "NFT", "meme stocks"],

)

profile = engine.get_profile()

check("Profile name set", profile["name"] == "Pierre-Henry Soria")

check("Goals set", len(profile["goals"]) == 3)

check("Interests set", len(profile["interests"]) == 7)

check("Projects set", len(profile["current_projects"]) == 3)

# ── 2. Context: check what the memory knows ──────────────

section("2. CONTEXT MEMORY")

ctx = engine.memory.get_context_snippets()

check("Context snippets populated", len(ctx) > 0, f"{len(ctx)} snippets")

tokens = engine.memory.get_context_tokens()

check("Context tokens populated", len(tokens) > 10, f"{len(tokens)} tokens")

# ── 3. Pipeline: run full pipeline ──────────────────────

section("3. FULL PIPELINE RUN")

notes_before = engine.store.count

result = engine.run_pipeline()

notes_after = engine.store.count

check("Pipeline succeeded", result["success"])

for step in result["steps"]:

check(f"Step: {step['name']}", step["status"] == "success")

print(f" Notes: {notes_before} → {notes_after}")

# ── 4. Deduplication: re-running should NOT duplicate ───

section("4. DEDUPLICATION TEST")

result2 = engine.run_pipeline()

notes_after_rerun = engine.store.count

check("Pipeline re-run succeeded", result2["success"])

check(

"No duplicate notes added",

notes_after_rerun == notes_after,

f"before={notes_after}, after={notes_after_rerun}",

)

# ── 5. Tag inference: notes should have rich tags ───────

section("5. TAG INFERENCE")

all_tags = set()

for n in engine.list_notes():

all_tags.update(n.get("tags", []))

check("Multiple tag categories", len(all_tags) > 2, f"{len(all_tags)} unique tags: {sorted(all_tags)[:10]}")

check("AI tag present", "ai" in all_tags or "AI" in {t.lower() for t in all_tags})

# ── 6. Evaluate digest ──────────────────────────────────

section("6. EVALUATE DIGEST")

eval_result = engine.evaluate_digest()

check("Digest evaluated", "error" not in eval_result)

check("Articles found", eval_result.get("total_articles", 0) > 0, f"{eval_result.get('total_articles')} articles")

check("AI ratio > 0", eval_result.get("ai_article_ratio", 0) > 0, f"{eval_result.get('ai_article_ratio', 0):.1%}")

check(

"Project fit > 0", eval_result.get("project_fit_score", 0) > 0, f"{eval_result.get('project_fit_score', 0):.1%}"

)

fit_before = eval_result["project_fit_score"]

# ── 7. Focus brief ──────────────────────────────────────

section("7. GENERATE FOCUS BRIEF")

brief = engine.generate_focus_brief()

check("Brief generated", brief.get("date") is not None)

check("Focus items present", len(brief.get("focus_items", [])) > 0, f"{len(brief.get('focus_items', []))} items")

check("Digest quality included", brief.get("digest_quality") is not None)

# ── 8. Self-improvement: learn + context enrichment ─────

section("8. SELF-IMPROVEMENT LOOP")

# Mark focus items as read

for item in brief.get("focus_items", [])[:2]:

engine.focus.mark_read(item["title"])

# Record a reading with insight

engine.memory.record_read(

title="Context-Aware Retrieval for AI Systems",

url="https://example.com/context-retrieval",

score=0.9,

tags=["ai", "retrieval", "context"],

insight="Context windows improve retrieval by 40%.",

)

# Record another with AI-relevant insight

engine.memory.record_read(

title="Building AI Agent Architectures",

url="https://example.com/agent-arch",

score=0.85,

tags=["ai", "agents", "architecture"],

insight="Multi-agent systems with shared context outperform single agents.",

)

check("Reading entries recorded", len(engine.memory.history) >= 2)

# Re-evaluate — context should now be richer

ctx_after = engine.memory.get_context_snippets()

tokens_after = engine.memory.get_context_tokens()

check("Context snippets grew", len(ctx_after) > len(ctx), f"{len(ctx)} → {len(ctx_after)}")

check("Context tokens grew", len(tokens_after) > len(tokens), f"{len(tokens)} → {len(tokens_after)}")

eval_after = engine.evaluate_digest()

fit_after = eval_after["project_fit_score"]

check(

"Context coverage increased",

eval_after["context_keyword_coverage"] >= eval_result.get("context_keyword_coverage", 0),

f"{eval_result.get('context_keyword_coverage', 0):.3f} → {eval_after['context_keyword_coverage']:.3f}",

)

print(f" Project fit: {fit_before:.1%} → {fit_after:.1%}")

# ── 9. Spaced repetition ────────────────────────────────

section("9. SPACED REPETITION")

due = engine.due_for_review()

check("Due-for-review returns list", isinstance(due, list))

# Advance a review

advanced = engine.advance_review("Context-Aware Retrieval for AI Systems")

check("Review advanced", advanced is not None)

if advanced:

check("Review level incremented", advanced["review_level"] >= 1, f"level={advanced['review_level']}")

check("Next review scheduled", advanced["next_review"] != "")

# ── 10. System status ───────────────────────────────────

section("10. SYSTEM STATUS")

status = engine.status()

check("Version present", status.get("version") == "0.1.0")

check("Notes count > 0", status.get("notes_count", 0) > 0, f"{status.get('notes_count')} notes")

check("Profile loaded", status.get("profile_loaded") is True)

print(f" {json.dumps(status, indent=2)}")

# ── Summary ─────────────────────────────────────────────

section("RESULTS")

total = PASS + FAIL

print(f" {PASS}/{total} checks passed, {FAIL} failed")

if FAIL == 0:

print(" ALL CHECKS PASSED — CortexOS self-improvement loop verified ✓")

else:

print(f" WARNING: {FAIL} checks failed — review above output")

return 0 if FAIL == 0 else 1

if __name__ == "__main__":

sys.exit(main())

Sunbelt Computer Software

PL/B Language Development and Support

Navigation Menu

Repository navigation

test_full_loop.py

test_full_loop.py

Sunbelt Computer Software

PL/B Language Development and Support

FilesExpand file tree

test_full_loop.py

Latest commit

History

test_full_loop.py

File metadata and controls