gen_ch4_figs.py
gen_ch4_figs.py
#!/usr/bin/env python3
"""Generate all SVG illustrations for Chapter 4 (Tools).
Figures (12 total):
fig4-1: MCP protocol sequence diagram (concrete message payloads)
fig4-2: Sub-Agent context preparation (4 strategies with examples)
fig4-3: Event-driven architecture (real event sources & payloads)
fig4-4: Async event processing (cancellation/queued/parallel timing)
fig4-5: Exp 4.4 — Event-driven agent architecture
fig4-6: Sync-async model contradiction (training vs deployment)
fig4-7: Exp 4.5 — Async agent with interruption
fig4-8: Tool discovery hierarchy (server→tool matching)
fig4-9: KV cache optimization (system prompt stability)
fig4-10: Tool self-evolution pipeline (multi-stage)
fig4-11: Exp 4.7 — Self-evolving agent pipeline
fig4-12: Voyager learning cycle (curriculum + skill library)
"""
import os
import sys
sys.path.insert(0, os.path.dirname(__file__))
from svg_lib import (
SVG, COLORS, FONT, MONO, STROKE_W, CORNER_R, _escape,
FS_TITLE, FS_BODY, FS_SMALL, FS_TINY, FS_LABEL,
)
OUT = os.path.join(os.path.dirname(__file__), 'images')
def _pill(svg, x, y, w, h, label, fill='light', font_size=FS_SMALL, bold=False):
svg.rect(x, y, w, h, fill=fill, rx=h // 2)
c = 'white' if fill in ('dark', 'darker') else 'text'
svg.text(x + w / 2, y + h / 2, label, size=font_size, fill=c, bold=bold)
def _seq_msg(svg, x1, x2, y, label, note=None, dash=False, note_side='above'):
"""Draw a sequence diagram message arrow with label."""
svg.arrow(x1, y, x2, y, dash=dash)
mid = (x1 + x2) / 2
if note_side == 'above':
svg.text(mid, y - 12, label, size=FS_SMALL, bold=True)
else:
svg.text(mid, y + 18, label, size=FS_SMALL, bold=True)
if note:
ny = y + 18 if note_side == 'above' else y + 34
svg.text(mid, ny, note, size=FS_TINY, fill='text_light')
# ──────────────────────── fig4-1 ────────────────────────
def fig4_1():
"""MCP protocol sequence diagram (concrete message payloads)"""
w, h = 880, 620
svg = SVG(w, h)
svg.text(w / 2, 30, "MCP protocol interaction sequence", size=FS_TITLE, bold=True)
cl_x, sv_x = 200, 680
svg.box(cl_x - 80, 50, 160, 44, "MCP Client", fill='medium', bold=True)
svg.box(sv_x - 80, 50, 160, 44, "MCP Server", fill='medium', bold=True)
svg.line(cl_x, 94, cl_x, 600, color='dark', dash=True)
svg.line(sv_x, 94, sv_x, 600, color='dark', dash=True)
# 1 initialize
y = 130
svg.arrow(cl_x + 4, y, sv_x - 4, y)
svg.text((cl_x + sv_x) / 2, y - 14, "initialize", size=FS_BODY, bold=True)
svg.code_block(cl_x + 30, y + 6, 350, [
'{"method": "initialize",',
' "capabilities": {"tools": true}}',
], font_size=FS_TINY, line_h=18)
# 2 initialize response
y = 200
svg.arrow(sv_x - 4, y, cl_x + 4, y, dash=True)
svg.text((cl_x + sv_x) / 2, y - 14, "initialize response", size=FS_BODY, bold=True)
svg.code_block(cl_x + 30, y + 6, 350, [
'{"serverInfo": {"name": "weather-server"},',
' "capabilities": {"tools": {"listChanged":true}}}',
], font_size=FS_TINY, line_h=18)
# 3 tools/list
y = 280
svg.arrow(cl_x + 4, y, sv_x - 4, y)
svg.text((cl_x + sv_x) / 2, y - 14, "tools/list", size=FS_BODY, bold=True)
svg.code_block(cl_x + 30, y + 6, 350, [
'{"method": "tools/list"}',
], font_size=FS_TINY, line_h=18)
# 4 tools/list response
y = 340
svg.arrow(sv_x - 4, y, cl_x + 4, y, dash=True)
svg.text((cl_x + sv_x) / 2, y - 14, "tools/list response", size=FS_BODY, bold=True)
svg.code_block(cl_x + 10, y + 6, 400, [
'{"tools": [{"name": "get_weather",',
' "inputSchema": {"city": "string"}}]}',
], font_size=FS_TINY, line_h=18)
# 5 tools/call
y = 420
svg.arrow(cl_x + 4, y, sv_x - 4, y)
svg.text((cl_x + sv_x) / 2, y - 14, "tools/call", size=FS_BODY, bold=True)
svg.code_block(cl_x + 30, y + 6, 350, [
'{"method": "tools/call",',
' "params": {"name": "get_weather",',
' "arguments": {"city": "Beijing"}}}',
], font_size=FS_TINY, line_h=18)
# 6 tools/call response
y = 510
svg.arrow(sv_x - 4, y, cl_x + 4, y, dash=True)
svg.text((cl_x + sv_x) / 2, y - 14, "tools/call result", size=FS_BODY, bold=True)
svg.code_block(cl_x + 30, y + 6, 350, [
'{"content": [{"type": "text",',
' "text": "Beijing: 22°C, sunny"}]}',
], font_size=FS_TINY, line_h=18)
# Phase labels on the left
svg.text(50, 165, "① Handshake", size=FS_SMALL, bold=True, fill='text_light')
svg.text(50, 310, "② Discovery", size=FS_SMALL, bold=True, fill='text_light')
svg.text(50, 465, "③ Invocation", size=FS_SMALL, bold=True, fill='text_light')
svg.save(os.path.join(OUT, 'fig4-1.svg'))
# ──────────────────────── fig4-2 ────────────────────────
def fig4_2():
"""Sub-Agent context preparation (comparison of 4 strategies)"""
w, h = 880, 530
svg = SVG(w, h)
svg.text(w / 2, 30, "Sub-Agent context passing strategies", size=FS_TITLE, bold=True)
strategies = [
("Minimal passing", "dark",
'"Query the status of order number 12345"',
"Zero context → privacy and security"),
("Manual filtering and passing", "medium",
'"User region: US\\nSummary: Inquiry about refund"',
"Explicit selection → controllable"),
("Automatic trimming and passing", "light",
'"User info + last 3 rounds\\n+ relevant tool results"',
"Rule-driven → balanced"),
("LLM-generated context", "code_bg",
'"LLM analyzes trajectory\\n→ structured context object"',
"Most intelligent → 1 extra call"),
]
col_w = 190
gap = 18
start_x = (w - 4 * col_w - 3 * gap) / 2
# Main Agent at top
svg.box(w / 2 - 100, 55, 200, 44, "Main Agent", fill='medium', bold=True)
svg.text(w / 2, 118, "How to prepare context for Sub-Agent?", size=FS_SMALL, fill='text_light')
for i, (title, fill, example, note) in enumerate(strategies):
x = start_x + i * (col_w + gap)
top_y = 145
svg.arrow(w / 2, 99, x + col_w / 2, top_y - 2)
svg.rect(x, top_y, col_w, 36, fill=fill)
tc = 'white' if fill in ('dark', 'darker') else 'text'
svg.text(x + col_w / 2, top_y + 18, title, size=FS_SMALL, bold=True, fill=tc)
svg.rect(x, top_y + 46, col_w, 80, fill='code_bg', stroke='dark', rx=4)
for j, line in enumerate(example.split('\\n')):
svg.mono(x + 8, top_y + 70 + j * 20, line, size=FS_TINY)
svg.text(x + col_w / 2, top_y + 150, note, size=FS_TINY, fill='text_light')
svg.box(x + 15, top_y + 175, col_w - 30, 36, "Sub-Agent", fill='light', font_size=FS_SMALL)
# Bottom: decision guide
svg.line(30, 395, w - 30, 395, color='dark', dash=True)
svg.text(w / 2, 418, "Selection guide", size=FS_BODY, bold=True)
guides = [
("Simple high-frequency calls", "Check weather, calculator", "→ Minimal"),
("Medium complexity", "Data query, file processing", "→ Automatic trimming"),
("Complex tasks", "Generate report, customer service", "→ LLM generation"),
]
gx = 80
for label, example, rec in guides:
svg.rect(gx, 438, 230, 70, fill='light')
svg.text(gx + 115, 458, label, size=FS_SMALL, bold=True)
svg.text(gx + 115, 478, example, size=FS_TINY, fill='text_light')
svg.text(gx + 115, 498, rec, size=FS_SMALL, bold=True, fill='darker')
gx += 260
svg.save(os.path.join(OUT, 'fig4-2.svg'))
# ──────────────────────── fig4-3 ────────────────────────
def fig4_3():
"""Event-driven architecture (specific event source and payload)"""
w, h = 880, 540
svg = SVG(w, h)
svg.text(w / 2, 30, "Event-driven asynchronous Agent architecture", size=FS_TITLE, bold=True)
# Left: Event sources
sources = [
("Email", 'on_email_reply', '{"from":"alice@...",\n "subject":"Re:meeting"}'),
("Timer", 'on_timer_expire', '{"task_id":"daily_report",\n "scheduled":"09:00"}'),
("Webhook", 'on_webhook', '{"repo":"agent-lib",\n "event":"pr_merged"}'),
("User", 'on_user_message', '{"text":"Check tomorrow\'s weather for me\n"}'),
]
src_x, src_w = 20, 155
svg.text(src_x + src_w / 2, 65, "Event source", size=FS_BODY, bold=True)
for i, (name, event_type, payload) in enumerate(sources):
y = 85 + i * 110
svg.box(src_x, y, src_w, 40, name, fill='medium', bold=True, font_size=FS_SMALL)
svg.mono(src_x + 5, y + 56, event_type, size=FS_TINY)
for j, pl in enumerate(payload.split('\n')):
svg.mono(src_x + 5, y + 74 + j * 16, pl, size=11)
# Middle: Event queue
q_x, q_w = 215, 190
svg.text(q_x + q_w / 2, 65, "Event queue", size=FS_BODY, bold=True)
svg.rect(q_x, 85, q_w, 390, fill='white', stroke='border', dash=True)
queue_events = [
("user.input", "Priority: normal", 'light'),
("email.reply", "Priority: normal", 'light'),
("user.interrupt", "Priority: urgent!", 'dark'),
("timer.trigger", "Priority: normal", 'light'),
]
for i, (evt, pri, fill) in enumerate(queue_events):
ey = 105 + i * 85
svg.rect(q_x + 10, ey, q_w - 20, 60, fill=fill, rx=4)
tc = 'white' if fill in ('dark', 'darker') else 'text'
svg.text(q_x + q_w / 2, ey + 22, evt, size=FS_SMALL, bold=True, fill=tc)
svg.text(q_x + q_w / 2, ey + 44, pri, size=FS_TINY, fill='white' if fill == 'dark' else 'text_light')
# Arrows from sources to queue
for i in range(4):
sy = 105 + i * 110
svg.arrow(src_x + src_w + 2, sy, q_x - 2, 120 + i * 85)
# Right: Agent processing
ag_x = 450
svg.text(ag_x + 200, 65, "Agent processing flow", size=FS_BODY, bold=True)
svg.arrow(q_x + q_w + 2, 280, ag_x - 2, 280, label="Fetch event")
steps = [
("Router", "LLM determines urgency", 'medium'),
("Append to trace", "Structured event format", 'light'),
("LLM inference", "Observe → Think → Act", 'light'),
("Tool execution", "Async/sync dispatch", 'light'),
("Result handling", "Notify/respond/store", 'medium'),
]
step_w, step_h = 360, 50
for i, (title, desc, fill) in enumerate(steps):
sy = 110 + i * 80
svg.rect(ag_x, sy, step_w, step_h, fill=fill)
svg.text(ag_x + 18, sy + step_h / 2, title, size=FS_SMALL, bold=True, anchor='start')
svg.text(ag_x + step_w - 12, sy + step_h / 2, desc, size=FS_TINY, fill='text_light', anchor='end')
if i < len(steps) - 1:
svg.arrow(ag_x + step_w / 2, sy + step_h + 2, ag_x + step_w / 2, sy + 78)
# Feedback loop
svg.arrow_curved(ag_x + step_w, 450, ag_x + step_w, 130, curve=45, label="Loop", dash=True, color='dark')
svg.save(os.path.join(OUT, 'fig4-3.svg'))
# ──────────────────────── fig4-4 ────────────────────────
def fig4_4():
"""Async event handling: timing comparison of three strategies"""
w, h = 880, 580
svg = SVG(w, h)
svg.text(w / 2, 30, "Three strategies for event handling", size=FS_TITLE, bold=True)
lane_x = 130
lane_w = 720
tl_x0 = lane_x + 10
tl_w = lane_w - 20
def time_bar(y, x_start_pct, x_end_pct, fill, label, h_bar=28):
xs = tl_x0 + tl_w * x_start_pct
xe = tl_x0 + tl_w * x_end_pct
svg.rect(xs, y, xe - xs, h_bar, fill=fill, rx=4)
svg.text((xs + xe) / 2, y + h_bar / 2, label, size=FS_TINY,
fill='white' if fill in ('dark', 'darker') else 'text')
# Timeline header
svg.text(tl_x0 + tl_w * 0.25, 55, "t₁", size=FS_SMALL, fill='text_light')
svg.text(tl_x0 + tl_w * 0.50, 55, "t₂", size=FS_SMALL, fill='text_light')
svg.text(tl_x0 + tl_w * 0.75, 55, "t₃", size=FS_SMALL, fill='text_light')
# ── Lane 1: Cancellation ──
y1 = 80
svg.rect(lane_x, y1, lane_w, 140, fill='white', stroke='border', dash=True)
svg.text(lane_x / 2, y1 + 70, "Cancellation", size=FS_BODY, bold=True)
svg.text(lane_x / 2, y1 + 95, "(Urgent)", size=FS_SMALL, fill='text_light')
time_bar(y1 + 15, 0.0, 0.40, 'medium', 'LLM reasoning...')
svg.line(tl_x0 + tl_w * 0.40, y1 + 10, tl_x0 + tl_w * 0.40, y1 + 130, color='border', dash=True)
svg.text(tl_x0 + tl_w * 0.40, y1 + 10, "⚡ user.interrupt: \"Stop!\"", size=FS_TINY, bold=True)
time_bar(y1 + 15, 0.40, 0.45, 'dark', '×', h_bar=28)
time_bar(y1 + 55, 0.0, 0.35, 'light', 'Tool executing...')
time_bar(y1 + 55, 0.40, 0.45, 'dark', '×', h_bar=28)
time_bar(y1 + 95, 0.47, 1.0, 'medium', 'New LLM reasoning (including interrupt event + clear queue)')
# ── Lane 2: Queued ──
y2 = 240
svg.rect(lane_x, y2, lane_w, 140, fill='white', stroke='border', dash=True)
svg.text(lane_x / 2, y2 + 70, "Queue-based", size=FS_BODY, bold=True)
svg.text(lane_x / 2, y2 + 95, "(Normal)", size=FS_SMALL, fill='text_light')
time_bar(y2 + 15, 0.0, 0.15, 'medium', 'LLM', h_bar=24)
time_bar(y2 + 15, 0.18, 0.60, 'light', 'Tool execution (search_web)')
time_bar(y2 + 15, 0.63, 0.90, 'medium', 'LLM comprehensive processing')
svg.line(tl_x0 + tl_w * 0.35, y2 + 46, tl_x0 + tl_w * 0.35, y2 + 130, color='dark', dash=True)
svg.text(tl_x0 + tl_w * 0.35, y2 + 58, "user: \"Only look at the last 1 month\"", size=FS_TINY, fill='text_light')
_pill(svg, tl_x0 + tl_w * 0.30, y2 + 65, 150, 24, "Enqueue waiting", fill='light', font_size=FS_TINY)
time_bar(y2 + 100, 0.63, 0.68, 'dark', '', h_bar=20)
svg.text(tl_x0 + tl_w * 0.61, y2 + 110, "Batch append: tool.result + user input", size=FS_TINY, fill='text_light', anchor='end')
# ── Lane 3: Parallel ──
y3 = 400
svg.rect(lane_x, y3, lane_w, 140, fill='white', stroke='border', dash=True)
svg.text(lane_x / 2, y3 + 70, "Parallel", size=FS_BODY, bold=True)
svg.text(lane_x / 2, y3 + 95, "(Independent)", size=FS_SMALL, fill='text_light')
time_bar(y3 + 15, 0.0, 0.80, 'light', 'Main task: Data analysis (long execution)')
svg.line(tl_x0 + tl_w * 0.30, y3 + 50, tl_x0 + tl_w * 0.30, y3 + 130, color='dark', dash=True)
svg.text(tl_x0 + tl_w * 0.30, y3 + 58, "user: \"How is the weather today?\"", size=FS_TINY, fill='text_light')
time_bar(y3 + 70, 0.32, 0.50, 'medium', 'Parallel LLM', h_bar=24)
time_bar(y3 + 70, 0.52, 0.62, 'dark', 'Weather', h_bar=24)
svg.text(tl_x0 + tl_w * 0.635, y3 + 82, "→ Reply to user immediately", size=FS_TINY, fill='text_light', anchor='start')
svg.text(tl_x0 + tl_w * 0.50, y3 + 115, "Tag: [Parallel with main task]", size=FS_TINY, fill='text_light')
svg.save(os.path.join(OUT, 'fig4-4.svg'))
# ──────────────────────── fig4-5 ────────────────────────
def fig4_5():
"""Experiment 4.4: Event-driven Agent Architecture"""
w, h = 880, 480
svg = SVG(w, h)
svg.text(w / 2, 30, "Experiment 4.4: Event-driven Agent Architecture", size=FS_TITLE, bold=True)
# Event sources (left column)
src_data = [
("on_user_message", "Web/App"),
("on_email_reply", "Email system"),
("on_github_pr_update", "GitHub"),
("on_timer_expire", "Timer"),
("on_webhook_received", "Webhook"),
("on_resource_alert", "System alert"),
]
svg.text(85, 65, "External event source", size=FS_BODY, bold=True)
for i, (evt, src) in enumerate(src_data):
y = 82 + i * 58
svg.rect(10, y, 150, 44, fill='light')
svg.text(85, y + 16, src, size=FS_SMALL, bold=True)
svg.mono(15, y + 36, evt, size=11)
# FastAPI Server (center)
svg.rect(200, 80, 200, 390, fill='white', stroke='border', dash=True)
svg.text(300, 100, "FastAPI server", size=FS_BODY, bold=True)
svg.rect(215, 120, 170, 50, fill='medium')
svg.text(300, 137, "HTTP endpoint", size=FS_SMALL, bold=True)
svg.text(300, 157, "POST /events/{type}", size=FS_TINY, fill='text_light')
svg.rect(215, 190, 170, 50, fill='light')
svg.text(300, 207, "Event router", size=FS_SMALL, bold=True)
svg.text(300, 227, "LLM determines urgency", size=FS_TINY, fill='text_light')
svg.rect(215, 260, 170, 50, fill='light')
svg.text(300, 277, "Event queue", size=FS_SMALL, bold=True)
svg.text(300, 297, "Priority sorting", size=FS_TINY, fill='text_light')
svg.rect(215, 330, 170, 50, fill='light')
svg.text(300, 347, "Agent loop", size=FS_SMALL, bold=True)
svg.text(300, 367, "Fetch → Reason → Execute", size=FS_TINY, fill='text_light')
svg.rect(215, 400, 170, 50, fill='medium')
svg.text(300, 417, "Session management", size=FS_SMALL, bold=True)
svg.text(300, 437, "Multi-threaded context", size=FS_TINY, fill='text_light')
for i in range(4):
svg.arrow(300, 170 + i * 70, 300, 190 + i * 70)
for i in range(6):
svg.arrow(160, 104 + i * 58, 213, 145)
# MCP Tools (right)
svg.text(610, 65, "MCP tool server", size=FS_BODY, bold=True)
tools = [
("Perception tools", "search_web, read_file\nread_webpage, parse_image"),
("execution tool", "code_interpreter\nvirtual_terminal, write_file"),
("collaboration tool", "browser_use\nrequest_human_approval"),
("notification tool", "send_email, send_slack\nsend_im_notification"),
]
for i, (name, desc) in enumerate(tools):
y = 82 + i * 100
svg.rect(460, y, 250, 80, fill='light')
svg.text(585, y + 22, name, size=FS_SMALL, bold=True)
for j, line in enumerate(desc.split('\n')):
svg.mono(470, y + 48 + j * 18, line, size=12)
svg.arrow(400, 355, 458, 180)
svg.arrow(458, 260, 400, 355)
# Persistent store
svg.rect(740, 82, 130, 380, fill='code_bg', stroke='dark', rx=4)
svg.text(805, 115, "persistence layer", size=FS_SMALL, bold=True)
items = ["conversation history", "event log", "scheduled task", "tool status", "audit trail"]
for i, item in enumerate(items):
svg.text(805, 160 + i * 55, item, size=FS_SMALL)
svg.save(os.path.join(OUT, 'fig4-5.svg'))
# ──────────────────────── fig4-6 ────────────────────────
def fig4_6():
"""sync-async model contradiction"""
w, h = 880, 520
svg = SVG(w, h)
svg.text(w / 2, 30, "synchronous training paradigm vs asynchronous deployment reality", size=FS_TITLE, bold=True)
# Top half: training pattern
svg.rect(20, 55, w - 40, 195, fill='white', stroke='border', dash=True)
svg.text(60, 78, "training paradigm (strictly synchronous sequence)", size=FS_BODY, bold=True, anchor='start')
_pill(svg, w - 200, 64, 160, 28, "API mandatory constraint", fill='dark', font_size=FS_SMALL)
steps_train = [
("Observation", 'medium', "User: Check Beijing weather"),
("Thinking", 'light', "Need to call weather tool"),
("Action", 'medium', "get_weather(Beijing)"),
("Observation", 'light', "22°C, sunny"),
]
bw, bh, gap = 180, 55, 22
sx = (w - (4 * bw + 3 * gap)) / 2
for i, (phase, fill, content) in enumerate(steps_train):
x = sx + i * (bw + gap)
svg.rect(x, 100, bw, bh, fill=fill)
svg.text(x + bw / 2, 120, phase, size=FS_SMALL, bold=True)
svg.text(x + bw / 2, 142, content, size=FS_TINY, fill='text_light')
if i < 3:
svg.arrow(x + bw + 2, 128, x + bw + gap - 2, 128)
svg.rect(sx, 170, 4 * bw + 3 * gap, 30, fill='code_bg', stroke='dark', rx=4)
svg.mono(sx + 10, 185,
"tool_call → next must be tool_result, otherwise API error", size=FS_TINY)
# Separator
svg.line(20, 262, w - 20, 262, color='dark', dash=True)
svg.text(w / 2, 280, "contradiction", size=FS_BODY, bold=True, fill='darker')
# Bottom half: async reality
svg.rect(20, 295, w - 40, 210, fill='white', stroke='border', dash=True)
svg.text(60, 318, "deployment reality (asynchronous events interleaved)", size=FS_BODY, bold=True, anchor='start')
_pill(svg, w - 200, 304, 160, 28, "Format conflict!", fill='dark', font_size=FS_SMALL)
# Async timeline
items = [
("Assistant", 'medium', "tool_call:\nget_weather(Beijing)", 0.0, 0.20),
("Waiting...", 'code_bg', "Tool execution ~5s", 0.22, 0.50),
("User interrupts", 'dark', "\"No need, \ncheck Shanghai's \"", 0.40, 0.55),
("???", 'code_bg', "When will tool_result arrive? \nHow to ensure format?", 0.57, 0.78),
("placeholder", 'light', "[Tool still executing, \nprioritize interruption]", 0.80, 1.0),
]
tl_x0, tl_w = 50, w - 100
for role, fill, txt, t0, t1 in items:
x0 = tl_x0 + tl_w * t0
x1 = tl_x0 + tl_w * t1
svg.rect(x0, 340, x1 - x0, 50, fill=fill, rx=4)
tc = 'white' if fill in ('dark', 'darker') else 'text'
svg.text((x0 + x1) / 2, 355, role, size=FS_TINY, bold=True, fill=tc)
for j, tl in enumerate(txt.split('\n')):
svg.text((x0 + x1) / 2, 372 + j * 14, tl, size=11, fill=tc)
svg.rect(50, 410, w - 100, 40, fill='code_bg', stroke='dark', rx=4)
svg.mono(60, 430,
"Solution: placeholder fixes format + non-urgent events enqueued + only interrupt when truly urgent",
size=FS_TINY)
# Bottom insight
svg.rect(140, 465, w - 280, 40, fill='dark')
svg.text(w / 2, 485,
"Fundamental solution: next-generation models need to be trained via RL in asynchronous environments",
size=FS_SMALL, fill='white', bold=True)
svg.save(os.path.join(OUT, 'fig4-6.svg'))
# ──────────────────────── fig4-7 ────────────────────────
def fig4_7():
"""Experiment 4.5: Asynchronous Agent with Interruption Capability"""
w, h = 880, 520
svg = SVG(w, h)
svg.text(w / 2, 30, "Experiment 4.5: Asynchronous Agent Interruption and Recovery", size=FS_TITLE, bold=True)
# Timeline
tl_y, tl_h = 60, 440
tl_x0, tl_w = 120, 740
# Lanes
lanes = [
("Agent", 80),
("Tool A", 180),
("Tool B", 260),
("Tool C", 340),
("Trajectory", 420),
]
for name, y in lanes:
svg.text(55, y, name, size=FS_SMALL, bold=True)
svg.line(tl_x0, y, tl_x0 + tl_w, y, color='dark', dash=True)
def tbar(y, t0, t1, fill, label, h_bar=22):
xs = tl_x0 + tl_w * t0
xe = tl_x0 + tl_w * t1
svg.rect(xs, y - h_bar / 2, xe - xs, h_bar, fill=fill, rx=3)
tc = 'white' if fill in ('dark', 'darker') else 'text'
svg.text((xs + xe) / 2, y, label, size=11, fill=tc)
# Phase 1: Agent starts 3 tools
tbar(80, 0.0, 0.12, 'medium', 'LLM: Launch 3 tools')
# Tools running
tbar(180, 0.13, 0.45, 'light', 'Script A: 3% per second → 33s to complete')
tbar(260, 0.13, 0.70, 'light', 'Script B: 2% per second → 50s...')
tbar(340, 0.13, 0.90, 'code_bg', 'Script C: 1% per second → 100s...')
# Event: tool A completes
t_done = 0.45
svg.line(tl_x0 + tl_w * t_done, 70, tl_x0 + tl_w * t_done, 450, color='border', dash=True)
svg.text(tl_x0 + tl_w * t_done, 62, "A completed", size=FS_TINY, bold=True)
# Agent checks others
tbar(80, 0.46, 0.58, 'medium', 'Query B, C progress')
tbar(420, 0.46, 0.58, 'light', 'B≈66% C≈33%')
# Cancel C (< 50%)
t_cancel = 0.60
svg.line(tl_x0 + tl_w * t_cancel, 70, tl_x0 + tl_w * t_cancel, 450, color='dark', dash=True)
svg.text(tl_x0 + tl_w * t_cancel, 62, "Cancel C", size=FS_TINY, bold=True, fill='darker')
tbar(340, 0.60, 0.65, 'dark', '×')
# B finishes
t_b_done = 0.70
svg.line(tl_x0 + tl_w * t_b_done, 70, tl_x0 + tl_w * t_b_done, 450, color='border', dash=True)
svg.text(tl_x0 + tl_w * t_b_done, 62, "B completed", size=FS_TINY, bold=True)
# Agent generates report
tbar(80, 0.72, 0.95, 'medium', 'LLM: Integrate A+B results to generate report')
tbar(420, 0.72, 0.95, 'light', 'A result + B result + C cancellation record')
# Annotations
svg.rect(tl_x0, 460, tl_w, 40, fill='code_bg', stroke='dark', rx=4)
svg.mono(tl_x0 + 10, 480,
"Key: placeholder injection + async completion event + cancel_tool(task_id) API",
size=FS_TINY)
svg.save(os.path.join(OUT, 'fig4-7.svg'))
# ──────────────────────── fig4-8 ────────────────────────
def fig4_8():
"""Tool discovery hierarchy (server→tool matching)"""
w, h = 880, 540
svg = SVG(w, h)
svg.text(w / 2, 30, "Hierarchical tool matching", size=FS_TITLE, bold=True)
# Query at top
svg.rect(250, 55, 380, 44, fill='medium')
svg.text(440, 77, "Agent: \"I need to query GitHub repository contributor statistics\"", size=FS_SMALL, bold=True)
svg.arrow(440, 99, 440, 130)
# discover_tools
svg.rect(300, 132, 280, 44, fill='dark')
svg.text(440, 154, "discover_tools(natural language requirement)", size=FS_SMALL, fill='white', bold=True)
svg.arrow(440, 176, 440, 210)
# Layer 1: Server matching
svg.rect(20, 210, w - 40, 110, fill='white', stroke='border', dash=True)
svg.text(55, 233, "Layer 1: Server matching (semantic similarity)", size=FS_BODY, bold=True, anchor='start')
servers = [
("GitHub", 0.92, 'dark'),
("Weather", 0.15, 'light'),
("Finance", 0.23, 'light'),
("ArXiv", 0.18, 'light'),
("File System", 0.31, 'light'),
]
sx = 50
for name, score, fill in servers:
svg.rect(sx, 255, 145, 50, fill=fill)
tc = 'white' if fill in ('dark', 'darker') else 'text'
svg.text(sx + 72, 272, name, size=FS_SMALL, bold=True, fill=tc)
svg.text(sx + 72, 292, f"Similarity: {score:.2f}", size=FS_TINY, fill='white' if fill == 'dark' else 'text_light')
sx += 165
# Arrow to layer 2
svg.arrow(123, 305, 123, 345)
svg.text(175, 330, "Top-1 server", size=FS_SMALL, fill='text_light')
# Layer 2: Tool matching within server
svg.rect(20, 345, w - 40, 160, fill='white', stroke='border', dash=True)
svg.text(55, 368, "Layer 2: Tool matching (26 tools within GitHub server)", size=FS_BODY, bold=True, anchor='start')
tools = [
("search_repositories", 0.41, "Search repositories"),
("list_contributors", 0.89, "Contributor list"),
("get_repo_stats", 0.85, "Repository statistics"),
("create_issue", 0.12, "Create Issue"),
("get_commit_history", 0.67, "Commit history"),
]
tx = 30
for name, score, desc in tools:
is_top = score > 0.80
fill = 'dark' if is_top else 'light'
svg.rect(tx, 388, 155, 55, fill=fill)
tc = 'white' if is_top else 'text'
svg.mono(tx + 5, 406, name, size=11, fill=tc)
svg.text(tx + 78, 428, f"{score:.2f} | {desc}", size=11, fill='white' if is_top else 'text_light')
tx += 170
# Bottom: result
svg.rect(180, 468, 520, 30, fill='code_bg', stroke='dark', rx=4)
svg.mono(190, 483, "Return Top-3: list_contributors, get_repo_stats, get_commit_history", size=12)
svg.save(os.path.join(OUT, 'fig4-8.svg'))
# ──────────────────────── fig4-9 ────────────────────────
def fig4_9():
"""KV Cache Optimization (System Prompt Stability)"""
w, h = 880, 560
svg = SVG(w, h)
svg.text(w / 2, 30, "KV Cache Optimization for Dynamic Tool Loading", size=FS_TITLE, bold=True)
# Left: naive approach
left_x = 30
svg.text(220, 65, "Naive Approach (Cache Invalidation)", size=FS_BODY, bold=True)
blocks_naive = [
("System Prompt", 120, 'medium', "You are an AI assistant...\n+ All tool schemas", "~50K tokens"),
("User Message", 100, 'light', "Query NVDA stock price", ""),
("Assistant", 80, 'light', "tool_call: ...", ""),
]
ny = 85
for label, bh, fill, content, note in blocks_naive:
svg.rect(left_x, ny, 380, bh, fill=fill, rx=4)
svg.text(left_x + 190, ny + 22, label, size=FS_SMALL, bold=True)
for j, line in enumerate(content.split('\n')):
svg.text(left_x + 190, ny + 44 + j * 20, line, size=FS_TINY, fill='text_light')
if note:
svg.text(left_x + 360, ny + 22, note, size=FS_TINY, fill='darker', anchor='end')
ny += bh + 8
svg.rect(left_x, ny + 5, 380, 40, fill='dark')
svg.text(left_x + 190, ny + 25, "Every time a new tool is loaded → entire cache invalidated!", size=FS_SMALL, fill='white', bold=True)
# Right: optimized approach
right_x = 460
svg.text(660, 65, "Optimized Approach (Cache Stability)", size=FS_BODY, bold=True)
blocks_opt = [
("System Prompt (Fixed)", 75, 'medium',
"You are an AI assistant...\nRole + Rules + Base Tools",
"~2K tokens | KV Cache"),
("Agent Status Bar (Lightweight)", 45, 'code_bg',
"Available tools: web_search, get_weather...",
"~200 tokens"),
("User: discover_tools", 40, 'light',
'"I need to check stock price"',
""),
("Tool Result", 55, 'light',
"Return get_stock_quote schema",
"Tool definitions here"),
("User Message", 40, 'light',
"Query NVDA stock price",
""),
("Agent Status Bar (Updated)", 45, 'code_bg',
"+get_stock_quote added",
"~220 tokens"),
]
oy = 85
for label, bh, fill, content, note in blocks_opt:
svg.rect(right_x, oy, 400, bh, fill=fill, rx=4)
svg.text(right_x + 200, oy + 16, label, size=FS_SMALL, bold=True)
for j, line in enumerate(content.split('\n')):
svg.text(right_x + 200, oy + 32 + j * 16, line, size=FS_TINY, fill='text_light')
if note:
svg.text(right_x + 390, oy + 16, note, size=11, fill='darker', anchor='end')
oy += bh + 5
svg.rect(right_x, oy + 5, 400, 40, fill='medium')
svg.text(right_x + 200, oy + 25, "System Prompt unchanged → KV Cache fully reused", size=FS_SMALL, bold=True)
# Bottom comparison
svg.line(30, 475, w - 30, 475, color='dark', dash=True)
comps = [
("Cache Hit Rate", "~0% (invalidated on every tool change)", "~95% (only hint changes slightly)"),
("First Token Latency", "High (recompute 50K tokens each time)", "Low (incremental compute ~200 tokens)"),
]
cy = 495
svg.text(250, cy, "Comparison Dimension", size=FS_SMALL, bold=True)
svg.text(500, cy, "Naive Approach", size=FS_SMALL, bold=True)
svg.text(740, cy, "Optimized Approach", size=FS_SMALL, bold=True)
for metric, naive, opt in comps:
cy += 28
svg.text(250, cy, metric, size=FS_TINY)
svg.text(500, cy, naive, size=FS_TINY, fill='text_light')
svg.text(740, cy, opt, size=FS_TINY, fill='text_light')
svg.save(os.path.join(OUT, 'fig4-9.svg'))
# ──────────────────────── fig4-10 ────────────────────────
def fig4_10():
"""Tool Self-Evolution Pipeline (Multi-Stage)"""
w, h = 880, 500
svg = SVG(w, h)
svg.text(w / 2, 30, "Agent Self-Evolution: From Requirements to Tools", size=FS_TITLE, bold=True)
# Pipeline stages
stages = [
("① Requirement Identification", 'medium', [
"Task: YouTube Subtitle Extraction",
"Agent: Current tools insufficient",
"→ Initiate self-evolution",
]),
("② Web Search", 'light', [
"search: youtube transcript",
"python library",
"→ Found 3 candidate libraries",
]),
("③ GitHub Exploration", 'light', [
"Visit jdepoix/youtube-",
"transcript-api repository",
"→ Read README + Examples",
]),
("④ Learning and Testing", 'light', [
"code_interpreter test:",
"from youtube_transcript",
" _api import ...",
]),
("⑤ Tool Encapsulation", 'medium', [
"Create MCP tool:",
"get_youtube_transcript",
"(video_id) → text",
]),
]
stage_w, stage_h = 155, 145
gap = 12
total_w = len(stages) * stage_w + (len(stages) - 1) * gap
sx = (w - total_w) / 2
for i, (title, fill, details) in enumerate(stages):
x = sx + i * (stage_w + gap)
svg.rect(x, 60, stage_w, stage_h, fill=fill)
svg.text(x + stage_w / 2, 82, title, size=FS_SMALL, bold=True)
svg.line(x + 10, 94, x + stage_w - 10, 94, color='dark')
for j, line in enumerate(details):
svg.mono(x + 8, 114 + j * 20, line, size=11)
if i < len(stages) - 1:
svg.arrow(x + stage_w + 2, 60 + stage_h / 2, x + stage_w + gap - 2, 60 + stage_h / 2)
# Tool registry at bottom
svg.arrow(w / 2, 205, w / 2, 240)
svg.rect(120, 240, w - 240, 50, fill='dark')
svg.text(w / 2, 265, "⑥ Register to tool library → reuse directly in future", size=FS_BODY, fill='white', bold=True)
# Reuse scenario
svg.arrow(w / 2, 290, w / 2, 320)
svg.rect(60, 320, w - 120, 160, fill='white', stroke='border', dash=True)
svg.text(w / 2, 345, "Tool reuse: next time encountering similar task", size=FS_BODY, bold=True)
svg.rect(80, 365, 340, 50, fill='code_bg', stroke='dark', rx=4)
svg.mono(90, 382, "Agent: \"I need to extract YouTube subtitles\"", size=FS_TINY)
svg.mono(90, 400, "→ search_tools(\"youtube transcript\")", size=FS_TINY)
svg.arrow(420, 390, 460, 390)
svg.rect(460, 365, 330, 50, fill='light')
svg.text(625, 382, "Hit! get_youtube_transcript", size=FS_SMALL, bold=True)
svg.text(625, 402, "Skip search and creation, call directly", size=FS_TINY, fill='text_light')
svg.rect(200, 430, 480, 35, fill='medium')
svg.text(w / 2, 448, "Tool layer + Knowledge layer + Strategy layer → more proficient with use", size=FS_SMALL, bold=True)
svg.save(os.path.join(OUT, 'fig4-10.svg'))
# ──────────────────────── fig4-11 ────────────────────────
def fig4_11():
"""Experiment 4.7: Agent searches for tools on the web, self-evolves"""
w, h = 880, 480
svg = SVG(w, h)
svg.text(w / 2, 30, "Experiment 4.7: Self-evolving Agent pipeline", size=FS_TITLE, bold=True)
# Top: minimal base tools
svg.rect(30, 60, w - 60, 48, fill='medium')
svg.text(w / 2, 76, "Basic tools (minimal set)", size=FS_SMALL, bold=True)
base_tools = ["web_search", "read_webpage", "code_interpreter", "create_tool", "search_tools"]
btx = 65
for t in base_tools:
tw = len(t) * 8 + 20
_pill(svg, btx, 82, tw, 22, t, fill='dark', font_size=11, bold=True)
btx += tw + 10
# Task input
svg.arrow(w / 2, 108, w / 2, 135)
svg.rect(100, 135, w - 200, 45, fill='code_bg', stroke='dark', rx=4)
svg.mono(110, 150,
"Task: \"NVDA latest stock price? Change from a week ago?\" → Agent: No financial tools!",
size=FS_TINY)
svg.mono(110, 168,
"→ Identify capability gap → Initiate self-evolution",
size=FS_TINY)
# Evolution pipeline
svg.arrow(w / 2, 180, w / 2, 210)
pipe_y = 210
pipe_stages = [
("web_search", "Search candidate solutions", 'light',
["\"python stock price API\"",
"→ yfinance, Alpha Vantage..."]),
("read_webpage", "Evaluate solutions", 'light',
["yfinance: free, no API key required",
"Alpha Vantage: requires registration..."]),
("code_interpreter", "Test and verify", 'light',
["import yfinance as yf",
"yf.Ticker('NVDA').history()"]),
("create_tool", "Encapsulate and register", 'medium',
["name: get_stock_data",
"schema: {ticker, period}"]),
]
pw = 190
pgap = 15
total_pw = len(pipe_stages) * pw + (len(pipe_stages) - 1) * pgap
px = (w - total_pw) / 2
for i, (tool, desc, fill, details) in enumerate(pipe_stages):
svg.rect(px, pipe_y, pw, 120, fill=fill)
_pill(svg, px + 10, pipe_y + 8, pw - 20, 22, tool, fill='dark', font_size=11, bold=True)
svg.text(px + pw / 2, pipe_y + 48, desc, size=FS_SMALL, bold=True)
for j, line in enumerate(details):
svg.mono(px + 8, pipe_y + 70 + j * 18, line, size=11)
if i < len(pipe_stages) - 1:
svg.arrow(px + pw + 2, pipe_y + 60, px + pw + pgap - 2, pipe_y + 60)
px += pw + pgap
# Tool registry
svg.arrow(w / 2, 330, w / 2, 360)
svg.rect(200, 360, w - 400, 44, fill='dark')
svg.text(w / 2, 382, "Tool library: get_stock_data registered", size=FS_BODY, fill='white', bold=True)
# Reuse
svg.arrow(w / 2, 404, w / 2, 430)
svg.rect(100, 430, w - 200, 40, fill='code_bg', stroke='dark', rx=4)
svg.mono(110, 442,
"Reuse verification: \"Query TSLA stock price\" → search_tools hit → directly call get_stock_data",
size=FS_TINY)
svg.mono(110, 458,
"Skip search/evaluation/testing phases → cost reduction 90%+",
size=FS_TINY)
svg.save(os.path.join(OUT, 'fig4-11.svg'))
# ──────────────────────── fig4-12 (Voyager, was fig4_voyager) ────────
def fig4_12():
"""Voyager learning loop (curriculum + skill library + iterative prompting)"""
w, h = 880, 520
svg = SVG(w, h)
svg.text(w / 2, 30, "Voyager: Agent architecture for continuous learning", size=FS_TITLE, bold=True)
svg.rect(20, 65, 260, 180, fill='white', stroke='border', dash=True)
svg.text(150, 88, "Automatic curriculum generator", size=FS_BODY, bold=True)
curriculum = [
"Input: current state + existing skills",
"Output: next exploration goal",
"",
"Example goal sequence:",
" Chop tree → Craft wooden planks",
" → Craft wooden pickaxe → Mine stone",
" → Craft furnace → Smelt iron ingots",
]
for i, line in enumerate(curriculum):
svg.mono(32, 112 + i * 20, line, size=12)
svg.rect(600, 65, 260, 180, fill='white', stroke='border', dash=True)
svg.text(730, 88, "Iterative prompting mechanism", size=FS_BODY, bold=True)
iterative = [
"Collect feedback on failure:",
" - Environment observation (error message)",
" - Self-verification result",
"",
"Integrate into LLM Prompt",
"→ Guide code improvement",
"→ Iterate multiple times until success",
]
for i, line in enumerate(iterative):
svg.mono(612, 112 + i * 20, line, size=12)
svg.arrow(280, 155, 370, 155, label="Goal")
svg.arrow(560, 155, 600, 155, label="Feedback")
svg.rect(370, 110, 190, 80, fill='medium')
svg.text(465, 140, "Agent execution", size=FS_BODY, bold=True)
svg.text(465, 165, "GPT-4 code generation", size=FS_SMALL, fill='text_light')
svg.arrow(465, 190, 465, 260)
svg.text(510, 230, "Success → Extract", size=FS_SMALL, fill='text_light')
svg.rect(120, 260, 640, 240, fill='white', stroke='border', dash=True)
svg.text(440, 283, "Skill Library — core of externalized learning", size=FS_BODY, bold=True)
skills = [
("chopTree()", "Chop tree\nBasic skill", "function chopTree() {\n bot.dig(nearest('log'));\n}"),
("craftPlanks()", "Craft wooden planks\nCall chopTree", "function craftPlanks() {\n chopTree(); craft('planks');\n}"),
("craftPickaxe()", "Craft wooden pickaxe\nCombine multiple skills", "function craftPickaxe() {\n craftPlanks(); craft('stick');\n craft('wooden_pickaxe');\n}"),
]
skx = 140
for name, desc, code in skills:
svg.rect(skx, 305, 190, 175, fill='light')
svg.text(skx + 95, 325, name, size=FS_SMALL, bold=True)
for j, dl in enumerate(desc.split('\n')):
svg.text(skx + 95, 347 + j * 18, dl, size=FS_TINY, fill='text_light')
svg.rect(skx + 10, 385, 170, 80, fill='code_bg', stroke='dark', rx=4)
for j, cl in enumerate(code.split('\n')):
svg.mono(skx + 18, 400 + j * 18, cl, size=11)
skx += 215
svg.arrow_curved(120, 380, 150, 245, curve=60, label="Existing skills", dash=True, color='dark')
svg.save(os.path.join(OUT, 'fig4-12.svg'))
# ──────────────────────── main ────────────────────────
def main():
os.makedirs(OUT, exist_ok=True)
figs = [
fig4_1, fig4_2, fig4_3, fig4_4, fig4_5, fig4_6,
fig4_7, fig4_8, fig4_9, fig4_10, fig4_11, fig4_12,
]
# Note: fig4_11 = Exp 4.7 self-evolving agent, fig4_12 = Voyager
# (ordered by chapter appearance)
for fn in figs:
fn()
print(f" ✓ {fn.__name__}: {fn.__doc__}")
print(f"\nGenerated {len(figs)} figures in {OUT}/")
if __name__ == '__main__':
main()