What does this skill do, and when should you use it?
This is one skill inside olaxbt/ai-market-maker (AIMM, an agentic AI hedge fund OS), located at src/agents/operator/2.1_pattern_recognition_bot/. It reads internal OHLCV bars and pre-computed Kalman filter estimates, runs the pattern detection logic from src/ta/pattern_engine.py, and returns a pattern label, setup score, support level and reasoning. It can be used programmatically via analyze(ticker, market_data) (returning a Tier-0 contract), through direct natural-language queries, or via a query endpoint like /pattern_recognition_bot?ticker=BTC/USDT. It only identifies and scores — it does not execute trades or manage positions. The repository is AGPL-3.0 licensed, which carries source-disclosure obligations for networked commercial deployments.
Accepts a ticker and market data, then calls pattern_engine.py for pattern detection; reads internal OHLCV bars and pre-computed Kalman filter estimates; answers direct queries such as "What pattern is forming on BTC?" or "Where is the support level?"; returns setup score, pattern label, support level and reasoning via /pattern_recognition_bot?ticker=BTC/USDT; exposes analyze(ticker, market_data), which outputs a structured Tier-0 contract.
- A crypto day trader who wants a quick read on which technical pattern is forming and where key support sits before entering a position.
- A quant researcher building a multi-agent pipeline who needs a standardized pattern-recognition node emitting a Tier-0 contract consumable by downstream risk or portfolio modules.
- An AIMM platform user who wants to pair this skill with the Technical TA Engine desk to diagnose patterns on pairs like BTC/USDT.
- An individual investor who wants to know a symbol's support level without manually inspecting charts.
How do you install this skill?
- Static review only; no code was executed — all reliability conclusions derive from source reading.
- The skill doc declares no permission boundaries, failure feedback, or rollback; inspect src/ta/pattern_engine.py behavior yourself before use.
- Setup scores and pattern labels are not independently validated and should not be used directly as trading signals; the repo itself notes a single profitable backtest is not proof of edge.
- The agentic path depends on overseas LLM services (OpenAI / Atlas Cloud) that may be unreachable from mainland China.
- Publisher identity is unverified; maintenance ownership and update path are only implicit at the repository level.
- Shell / CLI
- Network access
- Local filesystem
Python 3.11+TA-Lib
The skill ships inside the ai-market-maker repository. Install the repo: git clone https://github.com/olaxbt/ai-market-maker.git && cd ai-market-maker, install uv (pip install uv), install the TA-Lib C library first (conda recommended: conda install -y ta-lib -c conda-forge), then run uv sync --extra dev. Alternatively install via OpenClaw: claw install https://github.com/olaxbt/ai-market-maker. The skill file lives at src/agents/operator/2.1_pattern_recognition_bot/SKILL.md. The README does not document standalone installation for this sub-skill.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/olaxbt/ai-market-maker.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/src/agents/operator/2.1_pattern_recognition_bot" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- What pattern is forming on BTC?
- Where is the support level?
Three ways: 1) call analyze(ticker, market_data) programmatically to get a Tier-0 contract with pattern and setup score; 2) ask direct natural-language questions like "What pattern is forming on BTC?" or "Where is the support level?"; 3) query /pattern_recognition_bot?ticker=BTC/USDT to receive setup score, pattern label, support level and reasoning.
What are this skill's strengths and limitations?
- Standardized output (Tier-0 contract) that plugs cleanly into downstream risk and portfolio modules.
- Supports both natural-language queries and programmatic calls.
- Dependencies are explicit: pattern_engine.py and Kalman estimates; no external paid data feed (uses internal OHLCV).
- The parent repo ships a backtesting framework with buy-and-hold benchmarks to validate the broader system.
- Identification and scoring only — it cannot execute trades, size positions or veto risk (those belong to other modules/skills in the repo).
- The SKILL.md itself is minimal: no details on pattern algorithms, scoring methodology or known limitations.
- Relies on internal OHLCV data and pre-computed Kalman estimates; how to feed it data outside the main repo is undocumented.
- Backtest figures in the README are illustrative, and the author explicitly warns that a single profitable backtest is not proof of edge.
- AGPL-3.0 imposes extra source-offer obligations for networked commercial use.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Pattern Recognition Bot (2.1) this page | 42 · Not recommended | ★ 2.1k | 2mo ago | AGPL-3.0 |
| Technical TA Engine | 42 · Not recommended | ★ 2.1k | 2mo ago | AGPL-3.0 |
| TradingView MCP AI Trading Intelligence | 50 · Use with care | ★ 5.1k | 11d ago | MIT |
| News Narrative Miner | 46 · Use with care | ★ 2.1k | 2mo ago | AGPL-3.0 |
| Lighter Agent Kit | 57 · Use with care | ★ 67 | 2mo ago | MIT |
How did FollowSkills review this skill?
Read-only analytical agent (patterns, support levels, setup scores) with no external side effects or credential needs; data flow (internal OHLCV + pre-computed Kalman) is disclosed; however the SKILL.md itself omits permission boundaries, confirmation mechanisms, or rollback paths — deducted.
Output contract (tier0/v1 JSON) is consistent with persona few-shots; repo includes Tier-0 consensus tests and deterministic fixtures covering pattern fields (setup_confidence_score, kalman_support), but no skill-specific tests for src/ta/pattern_engine.py or the query endpoint, and no described failure feedback on abnormal input — deducted.
Scenario is clear (crypto chart pattern and support queries) with explicit trigger examples; but no non-fit boundaries (non-crypto assets, sparse data), English only, and no disclosure that core function depends on overseas LLM services (OpenAI/Atlas) that may be unreachable from mainland China — deducted.
Docs are readable with a clear contract and listed dependency paths; but no skill-level versioning/changelog, no FAQ or known-limitations section, maintenance responsibility only implicit at repo level; AGPL-3.0 license is clear at repo level — deducted.
README claims the pattern desk participates in weighted arbitration with backtest results, but output usability and numeric quality (setup score) cannot be statically verified, and marginal value over directly using TA-Lib lacks evidence — deducted.
CI workflow, pytest suite, and deterministic fixtures covering pattern-agent fields provide partially auditable material; but no targeted test evidence for this skill's key paths and no independent third-party reproduction — deducted.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →