Budget: 4000 UAH Deadline: 2 days
I am a system architect and I am interested in implementing this project with you. My proposal: I will do everything strictly according to your technical specifications, but additionally I will implement my filtering algorithm. For you, this will not cost a penny extra, but in the end, the system will work 2-3 times more efficiently than any standard solution. Essentially, you get what you ordered, plus a "safety cushion" that filters out false inputs and preserves the deposit. You will feel the benefit immediately over time.
Viktor N.
Winning proposal- Projects 24
- Rating 4.8
- Rating 2 517
Budget: 3235 UAH Deadline: 1 day
Good day. I am ready to implement the indicator.
I have extensive experience in writing indicators for both TV and custom ones in Python.
Right now, I am working on similar tasks - I can show them in private messages or during a screen demonstration in a call.
Proposals concealed
Proposals are currently absent
Budget: 20000 UAH Deadline: 30 days
Good day! I write indicators, trading experts in different languages. Contact me, we will discuss the details.
-
Pavel Shtemenko 17 JanuaryЗдравствуйте. С v5 дел имел не мало. Потому вопрос, а криптой вы сможете оплатить? У вас в коде кстати не рассмотрен вариант когда нет ни одной выбранной биржи
-
Ruslan P. 17 JanuaryСкриптов открытого интереса вагон. Или что именно вы хотите отследить? Аномалию?
-
Holovatyi Mykola 18 JanuaryПин ниже. Могу ускорить интеграцию в 3–5 раз, + устойчивость системы до х2 за счёт стандартизированного API и контроля режима.
//@version=6
indicator("AIF-ilter — Guard Controller PRO v6", overlay=false, max_labels_count=500)
// ======================================================
// Add-on: AUTO profile (switches Conservative/Balanced/Aggressive by volatility)
// - Volatility proxy: ATR% + |Z|
// - Auto picks profile each bar (and shows it in banner/table)
// ======================================================
// ===================== INPUTS ======================
grpMode = "Mode"
grpCore = "Core"
grpCtrl = "Controller"
grpGuard = "Guard / Kill-switch"
grpAuto = "Auto Profile"
grpViz = "Visual"
mode = input.string("Guard", "Mode", options=["Basic","Pro","Guard"], group=grpMode)
profileSel = input.string("Auto", "Profile", options=["Auto","Conservative","Balanced","Aggressive"], group=grpMode,
tooltip="Auto: профиль выбирается по волатильности (ATR% + |Z|)")
src = input.source(close, "Source", group=grpCore)
lenMA = input.int(55, "Trend MA", minval=1, group=grpCore)
lenMom = input.int(10, "Momentum (ROC)", minval=1, group=grpCtrl)
lenZ = input.int(80, "Stability window", minval=20, group=grpCtrl)
lenATR = input.int(14, "ATR length", minval=1, group=grpCtrl)
baseThr = input.float(65.0, "Base threshold", minval=1, maxval=99, step=0.5, group=grpCtrl)
adaptive = input.bool(true, "Adaptive threshold", group=grpCtrl)
biasLong = input.float(0.0, "Long bias", step=0.5, group=grpCtrl)
biasShort= input.float(0.0, "Short bias", step=0.5, group=grpCtrl)
// Auto profile cutoffs (loose by default)
atrLow = input.float(0.8, "ATR% low", step=0.1, group=grpAuto, tooltip="Ниже — можно Aggressive")
atrHigh = input.float(1.8, "ATR% high", step=0.1, group=grpAuto, tooltip="Выше — Conservative")
zHigh = input.float(2.5, "|Z| high", step=0.1, group=grpAuto, tooltip="Если |Z| выше — Conservative")
useGuard = input.bool(true, "Enable guard", group=grpGuard)
gapPctIn = input.float(1.2, "Gap block (%)", step=0.1, group=grpGuard)
spikeZIn = input.float(4.0, "Spike block (Z)", step=0.5, group=grpGuard)
freezeIn = input.int(3, "Freeze bars after guard", minval=0, group=grpGuard)
showBands = input.bool(true, "Show thresholds", group=grpViz)
showMarks = input.bool(true, "Show marks", group=grpViz)
showTable = input.bool(true, "Show table", group=grpViz)
showBanner = input.bool(true, "Show status banner", group=grpViz)
heatmap = input.bool(true, "Heatmap background", group=grpViz)
heatStr = input.int(85, "Heat intensity (0..100)", minval=0, maxval=100, group=grpViz)
// ===================== HELPERS ======================
clamp(x, lo, hi) => math.max(lo, math.min(hi, x))
pct(x) => str.tostring(x, "#.00") + "%"
// ===================== MODE FLAGS ======================
isBasic = mode == "Basic"
isGuard = mode == "Guard"
guardEnabled = (isGuard and useGuard)
// ===================== CORE =========================
c = src
ma = ta.ema(c, lenMA)
trendUp = c > ma
trendDn = c < ma
// ===================== VOL METRICS (for AUTO) =========
atr = ta.atr(lenATR)
atrPct = c != 0 ? (atr / c) * 100.0 : 0.0
// Z-score of 1-bar returns (stability)
ret1 = nz(c[1]) != 0 ? (c - c[1]) / c[1] * 100.0 : 0.0
mu = ta.sma(ret1, lenZ)
sd = ta.stdev(ret1, lenZ)
z = sd != 0 ? (ret1 - mu) / sd : 0.0
absZ = math.abs(z)
// ===================== PICK PROFILE ===================
// Auto logic:
// - If very stable (ATR% low AND |Z| low): Aggressive
// - If very unstable (ATR% high OR |Z| high): Conservative
// - Else: Balanced
string prof =
profileSel != "Auto" ? profileSel :
(atrPct <= atrLow and absZ <= (zHigh * 0.6)) ? "Aggressive" :
(atrPct >= atrHigh or absZ >= zHigh) ? "Conservative" :
"Balanced"
// ===================== PROFILE PARAMS =================
// Threshold, gains, guard sensitivity
thrAdj =
prof == "Conservative" ? 8.0 :
prof == "Aggressive" ? -6.0 : 0.0
momGain =
prof == "Conservative" ? 5.0 :
prof == "Aggressive" ? 7.5 : 6.0
stabPenalty =
prof == "Conservative" ? 14.0 :
prof == "Aggressive" ? 9.0 : 12.0
gapPct =
prof == "Conservative" ? gapPctIn * 0.85 :
prof == "Aggressive" ? gapPctIn * 1.25 : gapPctIn
spikeZ =
prof == "Conservative" ? spikeZIn * 0.85 :
prof == "Aggressive" ? spikeZIn * 1.25 : spikeZIn
freezeN =
prof == "Conservative" ? freezeIn + 2 :
prof == "Aggressive" ? math.max(freezeIn - 1, 0) : freezeIn
// ===================== CONTROLLER ======================
// Momentum
mom = ta.roc(c, lenMom)
scoreMom = clamp(50.0 + mom * momGain, 0.0, 100.0)
// Stability score
scoreStab = clamp(100.0 - absZ * stabPenalty, 0.0, 100.0)
// Trend score
scoreTrend = trendUp ? 100.0 : trendDn ? 0.0 : 50.0
// Confidence fusion
confBase = 0.45*scoreTrend + 0.35*scoreMom + 0.20*scoreStab
confLong = clamp(confBase + biasLong, 0.0, 100.0)
confShort = clamp((100.0 - confBase) + biasShort, 0.0, 100.0)
// Threshold (optional adaptive by |Z|)
thrBase = clamp(baseThr + thrAdj, 50.0, 95.0)
thr = adaptive ? clamp(thrBase + absZ * 3.0, 50.0, 95.0) : thrBase
upper = thr
lower = 100.0 - thr
// ===================== GUARD / KILL SWITCH ==========
gap = nz(c[1]) != 0 ? math.abs((open - c[1]) / c[1]) * 100.0 : 0.0
gapBad = gap > gapPct
spikeBad = absZ > spikeZ
naBad = na(c) or na(open)
guardHit = guardEnabled and (gapBad or spikeBad or naBad)
var int freeze = 0
if barstate.isnew
freeze := math.max(freeze - 1, 0)
if guardHit
freeze := freezeN
guardOK = (not guardEnabled) or (freeze == 0 and not guardHit)
guardReason =
not guardEnabled ? "OFF" :
naBad ? "NA" :
gapBad ? "GAP" :
spikeBad ? "SPIKE" :
guardHit ? "BLOCK" :
(freeze > 0 ? "FREEZE" : "OK")
// ===================== SIGNALS =======================
longOk = (not isBasic) and guardOK and trendUp and confLong >= upper
shortOk = (not isBasic) and guardOK and trendDn and confShort >= upper
longEdge = longOk and not longOk[1]
shortEdge = shortOk and not shortOk[1]
// ===================== HEATMAP =======================
heatAlpha = clamp(100 - heatStr, 0, 100)
isHotLong = confLong >= upper and trendUp
isHotShort = confShort >= upper and trendDn
bg =
not heatmap ? na :
not guardOK ? color.new(color.orange, 85) :
isHotLong ? color.new(color.lime, heatAlpha) :
isHotShort ? color.new(color.red, heatAlpha) :
color.new(color.gray, 95)
bgcolor(bg)
// ===================== PLOTS =========================
hline(50, "Mid", color=color.new(color.gray, 70))
plot(confBase, "Confidence", linewidth=2)
plot(showBands ? upper : na, "Thr (Long)", color=color.new(color.lime, 0))
plot(showBands ? lower : na, "Thr (Short)", color=color.new(color.red, 0))
plotshape(showMarks and (not isBasic) and longEdge, title="LONG", style=shape.triangleup, location=location.bottom, text="L", color=color.new(color.lime, 0), size=size.tiny)
plotshape(showMarks and (not isBasic) and shortEdge, title="SHORT", style=shape.triangledown, location=location.top, text="S", color=color.new(color.red, 0), size=size.tiny)
plotshape(showMarks and guardEnabled and guardHit, title="GUARD", style=shape.circle, location=location.top, text="G", color=color.new(color.orange, 0), size=size.tiny)
// Alerts
alertcondition(longEdge, title="AIF-ilter LONG", message="AIF-ilter LONG on {{ticker}} (profile=" + prof + ")")
alertcondition(shortEdge, title="AIF-ilter SHORT", message="AIF-ilter SHORT on {{ticker}} (profile=" + prof + ")")
alertcondition(guardHit, title="AIF-ilter GUARD", message="AIF-ilter GUARD (reason=" + guardReason + ", profile=" + prof + ") on {{ticker}}")
// ===================== STATUS BANNER ==================
var label banner = na
if showBanner and barstate.islast
label.delete(banner)
status =
guardEnabled and not guardOK ? "GUARD: " + guardReason :
longOk ? "READY: LONG" :
shortOk ? "READY: SHORT" :
"STANDBY"
statusColor =
guardEnabled and not guardOK ? color.new(color.orange, 0) :
longOk ? color.new(color.lime, 0) :
shortOk ? color.new(color.red, 0) :
color.new(color.silver, 0)
banner := label.new(bar_index, confBase,
"AIF-ilter PRO • " + mode + " • " + prof + " • " + status,
style=label.style_label_left,
textcolor=color.black,
color=statusColor)
// ===================== INFO TABLE ======================
var table t = table.new(position.top_right, 2, 12, frame_width=2, border_width=1,
bgcolor=color.new(color.black, 88))
if showTable and barstate.islast
table.cell(t, 0, 0, "Metric", text_color=color.white, bgcolor=color.new(color.blue, 70))
table.cell(t, 1, 0, "Value", text_color=color.white, bgcolor=color.new(color.blue, 70))
table.cell(t, 0, 1, "Mode")
table.cell(t, 1, 1, mode)
table.cell(t, 0, 2, "Profile")
table.cell(t, 1, 2, prof, text_color = prof == "Conservative" ? color.orange : prof == "Aggressive" ? color.lime : color.white)
table.cell(t, 0, 3, "Confidence")
table.cell(t, 1, 3, str.tostring(confBase, "#.0"))
table.cell(t, 0, 4, "Threshold")
table.cell(t, 1, 4, str.tostring(upper, "#.0") + (adaptive ? " (A)" : ""))
table.cell(t, 0, 5, "Trend")
tr = trendUp ? "UP" : trendDn ? "DOWN" : "FLAT"
table.cell(t, 1, 5, tr, text_color = trendUp ? color.lime : trendDn ? color.red : color.silver)
table.cell(t, 0, 6, "ATR%")
table.cell(t, 1, 6, pct(atrPct), text_color = atrPct >= atrHigh ? color.orange : atrPct <= atrLow ? color.lime : color.white)
table.cell(t, 0, 7, "ROC(" + str.tostring(lenMom) + ")")
table.cell(t, 1, 7, pct(mom), text_color = mom >= 0 ? color.lime : color.red)
table.cell(t, 0, 8, "|Z|")
table.cell(t, 1, 8, str.tostring(absZ, "#.00"), text_color = absZ >= zHigh ? color.orange : color.white)
table.cell(t, 0, 9, "Gap %")
table.cell(t, 1, 9, pct(gap), text_color = gapBad ? color.orange : color.white)
table.cell(t, 0, 10, "Guard")
table.cell(t, 1, 10, guardReason,
text_color = not guardEnabled ? color.gray : guardReason == "OK" ? color.lime : color.orange)
sig = longEdge ? "LONG" : shortEdge ? "SHORT" : "NONE"
table.cell(t, 0, 11, "Signal")
table.cell(t, 1, 11, sig, text_color = longEdge ? color.lime : shortEdge ? color.red : color.gray)
-
Holovatyi Mykola 18 JanuaryУже готово:
Отдельный сервис с API вместо встраивания в код, корректную работу с временем (dt) при лагах и джиттере, правильное хранение истории данных без искажений, фильтрацию шума на входе, устойчивое управление выходом и явные метрики состояния системы — сложность, управляющий сигнал, стабильность и ошибку. Теперь это можно подключать к любому источнику данных и использовать в продакшене без переписывания логики.
Могу доп: (патентная разработка)
Авторизацию и лимиты, алерты и мониторинг, автонастройку параметров под режим, дополнительные метрики сложности, масштабирование и деплой под Docker/Kubernetes.
Current freelance projects in the category Cryptocurrency & Blockchain
Aspect
Development of a Telegram Mini App (Bot + WebApp) for selling digital goods and in-game currencies (Python / aiogram) Project description: A Telegram bot with a full WebApp interface is required for the automated sale of digital goods (Telegram Stars, Telegram Premium, in-game currencies/donations for PUBG Mobile, Steam, etc.). The project is called Aspect App. The main focus is on speed, smooth UI/UX in a Web3/minimalist style, and complete automation of product delivery after payment. Technology stack: Backend: Python 3.10+, aiogram 3.x (FastAPI for Webhook/API is welcome). Database: PostgreSQL / Redis (for sessions and caching). Frontend (WebApp): React.js / Vue.js / Next.js (at the developer's discretion, speed and smoothness of animations are important). Integrations: Telegram WebApp API. Main functionality (MVP): 1. Client side (Telegram Mini App): Main screen: Product categories (Telegram Assets, In-game Donations). Banner grid for promotions. Product catalog: Product cards with volume selection (e.g., 50, 100, 500 Telegram Stars or UC). Fields for entering player data (Player ID for PUBG). Cart and Checkout: Quick purchase in 2 clicks. Payment module: Integration of payment methods (CryptoBot API / TON Connect / other fiat gateways by agreement). Personal account: Order history, status tracking, referral system (balance within the app). 2. Admin panel (built into WebApp or separate bot): Catalog management (adding/removing products, changing prices). Order monitoring and sales statistics. User database mailing system. 3. Automation logic (A plus): Readiness of the architecture for integration with supplier APIs (API for auto-purchasing Stars/UC). At the MVP stage, some products may be issued with codes from the database. What is required from the contractor: 1. Development of the database and backend architecture in Python. 2. Layout and integration of WebApp (design mockup to be discussed, clean Web3 style, Glassmorphism is important). 3. Setting up secure transactions and payment webhooks. 4. Deployment on the server (Docker, Ubuntu, SSL setup). Candidate requirements: Experience in commercial development of Telegram WebApps for at least 1 year. Portfolio with launched and working Mini Apps (please provide links in your response). Clean, documented code. Working conditions: Work only through the Secure Deal (Safe / Escrow) of the platform. Payment is staged (the project is divided into Sprint 1: Backend + DB, Sprint 2: Frontend WebApp, Sprint 3: Payment integration and testing). Budget: To be discussed with the successful candidate (please indicate your price range and timelines in your response)
Looking for a streamer, ideally with subscribers on Twitch; the more subscribers, the better. We need to promote the website so that as many people as possible learn about it.
We are looking for a specialist in Solana security programs to conduct an audit of the smart contract before deployment on the mainnet. Contract stack: Rust + Anchor framework SPL Token (transfer/storage of tokens in PDA) Mechanics: timing rounds, escrow of bets, resolver of final price, calculation and distribution of payouts, claim instructions, emergency pause + refund, fee configuration through admin authority Check for common vulnerabilities in Solana/Anchor: missing ownership/signer checks, account validation, integer overflow, PDA seeds collisions, reentrancy patterns, authority key privileges Analysis of payout distribution logic (rounding, edge cases with multiple participants) Verification of emergency mechanisms (pause, refund) — whether they can be bypassed or abused Report with classification of findings by criticality (Critical/High/Medium/Low) and recommendations Re-check after fixes (re-audit fixes) Requirements for the performer: Proven experience in auditing or developing Solana programs (Rust/Anchor) — please provide links to GitHub, previous audits, or examples of found vulnerabilities Understanding of SPL Token and PDA-escrow patterns Experience with static analysis tools (Soteria, Sec3, cargo-audit) will be a plus In your response, please indicate: Experience specifically with Solana (not EVM) — specific projects Estimated cost and timeline for a contract of this scope Report format (example, if available)
What we are calculating in the project: Realized PnL, ROI, and Win Rate for cryptocurrency wallets - how profitable the wallet traded a certain token over the selected period. Based on what data: the history of on-chain transactions of the wallet (swaps, token transfers) + the market price of the token at the time of each transaction. The main data source is Moralis: two calls during the initial loading of the wallet - native ETH transfers and all ERC-20 token transfers. What we are comparing with: Nansen.io - we take it as a benchmark, comparing our calculated metrics with what Nansen shows for the same wallets over the same period. Problem: our figures significantly differ from Nansen, and we do not fully understand the rules by which some actions of the wallet should be classified for PnL purposes. We need to fix the calculation of Realized PnL, ROI, and Win Rate to match Nansen. In your application, please write: - experience with similar tasks - experience with Nansen - experience with Moralis - experience with DeFiLlama - experience with articles on crypto transactions - approximate cost and timeline for the fix
It is important to start - TODAY I need a person who has experience in writing trading bots on DEXs (needed on the BSC network) with an understanding of transaction costs, gas, etc. I want to test several trading strategies on a real trading agent I need to quickly develop an autonomous trading agent that uses the Trust Wallet and CoinMarketCap APIs to execute trades on the BNB Chain (BSC) based on 3–5 embedded strategies. Technology stack (mandatory) - Trust Wallet Agent Kit (TWAK) — for self-custodial signing and executing transactions (https://portal.trustwallet.com) - CoinMarketCap AI Agent Hub — for obtaining market data and signals (https://coinmarketcap.com/api/agent) - BNB AI Agent SDK — for quick integration (https://github.com/bnb-chain/bnbagent-sdk) Functional requirements Data (CMC AI Agent Hub) - Data retrieval via MCP or x402 (https://coinmarketcap.com/api/agent) - Use of ready-made Skills for RSI, MACD, Fear & Greed, volumes Execution (Trust Wallet Agent Kit) - Local signing of transactions — keys remain with the user (https://portal.trustwallet.com) - Autonomous mode: the agent signs and sends trades by itself - Limitations: drawdown, daily limit, stop-loss, token whitelist - Use of x402 for payment of data/computations Strategies - Implement 3–5 described strategies with the ability to switch between them - Examples: combination of funding rates + Fear & Greed, DCA based on sentiments with filters Technical requirements - Language: Python (preferred) or TypeScript/JavaScript - Tools: TWAK CLI, CMC CLI, BNB AI Agent SDK Who can take this on?