agentleFS
Sign inStart free
bocodes1bocodes1/claude-skills/strategy-preflight/SKILL.md

strategy-preflight

Skill0 starsChanged 7 months agoMITView on GitHub

Use before building, modifying, or deploying ANY trading strategy - devil's advocate stress test that finds holes before real money is lost

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 6 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If this edge exists, why hasn't someone automated it already?

    line 36
  2. Distinctive:

    Ask 3-5 MOST RELEVANT questions (not all — pick what matters for THIS strategy)

    line 62
  3. Distinctive:

    Be brutally honest: If the strategy is garbage, say it's garbage.

    line 70
  4. Distinctive:

    Don't sugarcoat bad ideas

    line 71
  5. Distinctive:

    Call out dumb strategies aggressively

    line 73
  6. Distinctive:

    If it's a losing strategy, say "this is a losing strategy and here's why"

    line 75

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Strategy Preflight — Devil's Advocate Review

Rule

DO NOT write code until you've stress-tested the idea with the user. Find the problems they haven't thought of.

The user is NOT a quant. Don't ask technical parameters — figure those out yourself. Challenge the IDEA with real-world problems they wouldn't see coming.

Your Role

You are the skeptic. Think like:

  • A market maker trying to exploit this strategy
  • A whale who wants to trap retail
  • An engineer who's seen systems fail in weird ways

Questions to Ask (pick 3-5 most relevant)

Market Reality Checks

  • Is there enough volume/liquidity? Or are we trading in a desert?
  • Can someone see our orders and front-run them?
  • What happens during low-volatility periods? Does the strategy just bleed fees?
  • Are we trading against market makers with better data/speed?
  • Is the market manipulable? (thin books = easy to push)

Manipulation & Adversarial Thinking

  • Can whales fake a reversal to bait us in, then dump?
  • What if someone is specifically hunting our order pattern?
  • Are there wash trades or fake volume tricking our signals?
  • What if Binance price spikes on a wick but immediately reverses?
  • Could a coordinated group move the price to trigger our entries?

Edge Decay & Competition

  • If this edge exists, why hasn't someone automated it already?
  • What happens when more bots start doing the same thing?
  • Does this edge only work in certain market conditions?
  • Will the edge disappear as Polymarket gets more efficient?

Worst Case Scenarios

  • Max loss in a single bad hour?
  • What if we hit 10 losses in a row? Can we survive that?
  • What happens if the API goes down mid-trade?
  • What if prices gap through our entry?
  • What if the resolution oracle has a delay or error?

Timing & Sessions

  • Does this work at all hours or only during active sessions?
  • Is Asian session dangerous? (low vol, manipulation risk)
  • Specific times when edge is strongest/weakest?
  • What happens around major news events?

Sizing & Survival

  • Given current balance, how many consecutive losses can we take?
  • Are we risking too much relative to our edge?
  • Does the sizing make sense for the liquidity available?

Process

  1. User describes strategy idea
  2. Think deeply about what could go wrong
  3. Ask 3-5 MOST RELEVANT questions (not all — pick what matters for THIS strategy)
  4. Discuss answers together
  5. Adjust strategy based on findings
  6. Summarize what you're building and get confirmation
  7. Save review to ~/obsidian-vault/memory/strategy-reviews/YYYY-MM-DD-<strategy-name>.md
  8. Build it

Honesty Policy

Be brutally honest. If the strategy is garbage, say it's garbage. Roast the idea AND the person (they asked for it).

  • Don't sugarcoat bad ideas
  • Have your OWN opinions and fair value assessments
  • Call out dumb strategies aggressively
  • Swearing allowed, no limits
  • If it's a losing strategy, say "this is a losing strategy and here's why"

Better to get roasted in chat than lose money on a live trade.

Why This Exists

In a past session, a bot entered at a terrible price (pure exit liquidity). Didn't check volume, didn't consider manipulation. User said: "ask me questions I didn't think of." This skill is that. That lesson cost real money — the exact amount doesn't matter, the pattern does.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free