EdgeIQ Market Watch Bot

Crypto-first market intelligence and monitoring skill for building or operating a lightweight alerting bot that polls public Binance and Coinbase endpoints,...

Security Scan

The skill appears to do what it says (poll public Binance/Coinbase data, compute simple signals, and optionally send Telegram/WhatsApp alerts); nothing in the files indicates hidden exfiltration or unrelated capabilities, but review of the full script and any runtime env-file you pass is recommended.

Purpose & Capability

Name/description match the code and README: the script fetches public Binance and Coinbase endpoints, computes momentum/volatility/Bollinger/volume signals, and supports Telegram/Twilio alert delivery. The requested capabilities are proportionate to a market-watch bot.

Instruction Scope

SKILL.md instructs running scripts/market_watch.py and using a .env or CLI flags for credentials. That scope is expected, but it gives the agent permission to read whatever .env you point it to — which could contain unrelated secrets. Also the provided source file was truncated in the listing here, so I could not inspect the final delivery logic for any unexpected external endpoints; you should inspect the remainder of scripts/market_watch.py before running.

Install Mechanism

No install spec (instruction-only with a shipped Python script). That is lower risk than arbitrary downloads; the code runs from the skill folder and uses standard library urllib. No external packages or download URLs are present.

Credentials

Registry metadata lists no required env vars, which is consistent with optional delivery credentials. The README and SKILL.md document optional vars (TELEGRAM_BOT_TOKEN, TWILIO_* etc.) that are reasonable for alert delivery. Be aware the script will read any env-file you give it (load_env_file sets os.environ), so avoid passing a .env that contains unrelated/high-privilege secrets.

Persistence & Privilege

always:false and no special persistence requested. The skill does not request to modify other skills or agent settings; autonomous invocation is allowed by platform default and is not itself flagged.

Assessment

This skill appears coherent for a crypto market-monitoring bot. Before installing or running it: (1) review the remainder of scripts/market_watch.py (the provided listing was truncated) to confirm there are no unexpected external endpoints or data exfiltration paths; (2) run first in --dry-run mode and with a minimal .env containing only the delivery credentials you intend to use (or none) — do not point it at a system .env that contains unrelated secrets; (3) if you enable Telegram/WhatsApp delivery, create low-privilege tokens or disposable chat endpoints and limit their scope; (4) consider running the script in a sandbox or isolated environment if you cannot fully audit the code. If you want, provide the rest of scripts/market_watch.py and I will re-check the alert/delivery logic and any remaining code paths.

Build and operate a crypto market watch bot that checks public exchange data on a short cadence, scores simple technical conditions, and sends concise informational alerts.

Product positioning

Keep the bot framed as monitoring and market intelligence.

Never present outputs as guaranteed entries, exits, or financial advice.

When a suggested action is useful, keep it neutral:

• watch closely
• review long bias
• review short bias
• risk-off / wait
• high volatility - no action

Always include the disclaimer: Informational market monitoring only. Not financial advice.

Tiers

• Free — 1 watchlist, console output only
• Pro ($19/mo) — Telegram/WhatsApp alerts, multiple symbols, JSON export
• Bundle ($39/mo) — packaged with the broader EdgeIQ Labs bundle

Core capabilities

1. Pull public market data

Use Binance spot endpoints for ticker and klines.

Use Binance futures funding-rate endpoint when available.

Use Coinbase Exchange as fallback coverage for ticker and candles.

See references/api-notes.md for the endpoint map.

2. Analyze each symbol

Compute at least:

• momentum
• realized volatility from recent returns
• Bollinger bands
• unusual volume spike ratio
• funding rate when available

Summarize the result per symbol with:

• current price
• exchange source
• signal label
• trigger reasons
• suggested review action
• disclaimer

3. Alert only when conditions trigger

Support safe dry-run testing.

Send alerts through:

• Telegram Bot API
• Twilio-style WhatsApp messaging when credentials are present

Keep payloads short enough for chat surfaces.

Main script

Use scripts/market_watch.py as the one-shot runner.

Typical runs:

python3 scripts/market_watch.py --symbols BTCUSDT --dry-run
python3 scripts/market_watch.py --symbols BTCUSDT,ETHUSDT --json --dry-run
python3 scripts/market_watch.py --env-file .env

Configuration

Prefer .env or CLI flags for:

• watchlist symbols
• candle interval and candle count
• momentum threshold
• volatility threshold
• Bollinger settings
• volume spike threshold
• funding threshold
• delivery credentials

Use .env.example as the template.

Deliverables in this skill folder

• scripts/market_watch.py — main app
• .env.example — config template
• README.md — operator guide
• references/api-notes.md — endpoint notes

Legal disclaimer

All outputs from this skill must remain informational alerts only and must not be framed as financial advice.