Content
68%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A well-structured, highly actionable skill body built around a real script, a complete parameter reference, and a genuine one-level reference file. The main gaps are missing validation/feedback steps in the workflow (capped for batch operations), a requirements.txt mention with no such file in the bundle, and some inline time-sensitive regulatory detail that would fit better in the reference.
Suggestions
Add explicit validation checkpoints to the workflow: verify the input's schema_version before planning, handle the zero-actionable-orders case, and confirm both output files (JSON and Markdown) were written before presenting results.
Remove or fix the 'Install requirements.txt' instruction — no requirements.txt exists in the bundle — and use a skill-relative script path (scripts/plan_breakout_trades.py) instead of the install-prefix path.
Move the FINRA intraday-margin dates and broker-constraint detail into references/minervini_entry_rules.md (or a dedicated compliance note) so the time-sensitive regulatory text is loaded only when needed.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly lean — dense tables for gate criteria and CLI parameters, terse section prose, no beginner-concept padding. Not 5: execution modes are stated twice (intro parenthetical and Step 3) and time-sensitive regulatory dates ('effective 2026-06-04... phase-in allowed through 2027-10-20') are inlined in the body rather than delegated. Not 3: the padding is minor and localized. | 4 / 5 |
Actionability | Provides a copy-paste-ready CLI invocation with real arguments, a complete parameter table with defaults, and named output files. Not 5: 'Install requirements.txt before running' references a file that does not exist in the bundle, and the script path uses an install-location prefix ('skills/breakout-trade-planner/scripts/...') rather than a skill-relative path. | 4 / 5 |
Workflow Clarity | Three steps are clearly sequenced and Step 2 ('Review Output') acts as a soft checkpoint, but there is no validation or feedback loop: no input schema_version check, no error handling when zero candidates pass the gate, and no verification that output files were written. The batch-operation guideline caps this at 3. Not 4: checkpoints are implicit rather than explicit validate-and-retry steps. | 3 / 5 |
Progressive Disclosure | Well-organized sections with one clearly signaled, one-level-deep reference ('references/minervini_entry_rules.md' — verified to exist and to complement rather than duplicate the body), and scripts kept as internal implementation. Not 5: the inline Minervini Gate table and the broker-regulation detail partially duplicate what the reference file could carry, leaving the body denser than a pure overview. | 4 / 5 |
Total | 15 / 20 Passed |