Decision Grading Journal

    1

    Turn every decision — human or AI — into a calibrating dataset. Gradeable in advance, or it doesn't count.

    $29

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Decision Grading Journal

    Example session with this skill installed

    I need to record a 'no action' decision on a BTC/USD trade setup. The current price is 64000 but my slippage hurdle wasn't met. Create the journal entry.

    • Read your context and instructions
    • Compiled the decision grading journal

    I have generated the decision record for the BTC/USD 'no action' call. The entry is bound to the current state and includes the mandatory edge definitions to ensure this can be graded against future price action.

    Decision Record: BTC-USD-2026-05-12-01

    FieldValue
    BindingTS: 2026-05-12T14:30:05Z | State: H4-Trend-09 | Config: v2.1.4
    Executable SideBid: 63995.50 | Ask: 64004.50 (Ref: 64000.00)
    Edge DefinitionGross: +150bps | Net: +110bps | Hurdle: 120bps
    Haircut ListFees: 5bps | Slippage: 20bps | Adversity: 15bps
    DecisionNULL (No Action)
    Reason StringExpected net edge (110bps) sits below the stand-down hurdle (120bps). Rejected Long side due to insufficient volatility buffer in the current window.
    Pre-ReadIf price exceeds 65500 within 4h, the discipline was correct but the hurdle is too conservative. If price stays sub-64500, the edge definition is calibrated.
    Baseline FlagSimple MA-Cross would have taken this trade.

    Next steps

    1. Append this record to journals/trading-log-2026-Q2.md.
    2. Monitor the 4-hour window to settle the pre-read.
    3. Run audit-checklist.md at EOD to score the reason string integrity.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    A decision you can't grade in advance is a decision you can't learn from. This skill is the full contract: entries must carry binding metadata, executable prices (not reference prices), a pinned edge definition with haircut list, a stand-down hurdle, and a reason string that engages both sides of the edge table. Grading rules cover pre-registered reads, proper scoring vs the market/consensus, NULL grades for no-action calls, why one binary outcome carries near-zero calibration information, why measurement fixes improve gradeability (not P&L), and how to concede cleanly when a review contradicts the record. Includes the one-bit counterfactual flag that closes the reasoning-vs-baseline attribution gap with no new instrumentation, and the Goodhart warning for goal metrics that pressure the discipline that is the actual edge. What's inside: SKILL.md (the record contract and grading rules), reference/audit-checklist.md (per-record audit checklist, three-line grade template, contradiction-handling procedure), examples/worked-example.md (one decision record graded end to end).

    Requires real-time capture of executable prices and data snapshots. Cannot retroactively audit records missing the original binding state or reason strings.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 5 days ago

    Permissions

    Read Files
    Terminal / Shell
    Listed5 days ago

    What's inside

    Frequently Asked Questions

    This skill eliminates "hindsight bias" by enforcing a strict record contract that captures data snapshots and pre-registered outcomes before results are known, allowing for objective process auditing.

    It is framework-agnostic and works with any system that can output JSON or Markdown, making it compatible with Python trading bots, TypeScript operations tools, or manual terminal-based workflows.

    You receive the standardized decision-journaling schema, the pre-registration and grading logic, and a set of command-line tools to calculate calibration metrics like Brier scores.

    No, this skill requires real-time data snapshots; it cannot retroactively audit decisions where the specific executable terms and timestamps were not captured at the moment of the event.

    The skill includes a 'process vs. outcome' checklist that isolates environmental noise and slippage, allowing you to identify if a failure was due to a flawed strategy or simply a statistical outlier.

    New in Data & Databases