Content
63%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.
The body delivers a clear overview, a concrete run command, and well-organized scoring tables, and its reference file is real and properly signaled. Its weaknesses are redundancy (duplicated zone tables and inlined methodology the script and reference already handle) and a provenance section that references external skills and flags not resolvable within this bundle.
Suggestions
Move the Warning System, Momentum Smoothing, and Historical Confidence Indicator sections into references/uptrend_methodology.md, keeping only a one-line pointer, since the body already states the script handles scoring during regular execution.
Merge the '7-Level Zone Detail' table into the 'Scoring Zones' table (or drop it) to remove the duplicated zone breakdown, and consider compressing the Market Top Detector comparison and bilingual trigger list.
Clarify the 'Same-run provenance' section: either scope it to this skill's verified '--run-id' flag (which scripts/uptrend_analyzer.py supports) or move the external market-regime-daily/Exposure Coach handoff details to the skill that orchestrates those tools; also fix the run command path to match the bundle layout (scripts/uptrend_analyzer.py).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly compact tables, but it duplicates zone detail across 'Scoring Zones' and '7-Level Zone Detail', inlines scoring internals the script already applies (EMA(3) smoothing, 10v10 acceleration windows, warning penalty stacking), and spends tokens on a comparison table for a different skill, a bilingual trigger list, and a dense provenance section about external tools. | 3 / 5 |
Actionability | Execution guidance is concrete (a runnable python3 command, numbered script behavior, named output files, explicit reference-loading rules), but the command path assumes an install layout ('skills/uptrend-analyzer/scripts/...') rather than the bundle's 'scripts/' location, and the provenance section instructs use of external tools ('market-regime-daily producers', Exposure Coach's '--require-provenance ... --as-of' mode) without resolvable commands. | 4 / 5 |
Workflow Clarity | The two-phase sequence (run script, then present the report with specified highlights) is clear and the script self-validates (malformed primary dates stop report generation), but there is no error-recovery guidance for the multi-tool provenance handoff described in the final section, which is itself under-sequenced. | 4 / 5 |
Progressive Disclosure | The single reference (references/uptrend_methodology.md, verified to exist) is one level deep and well signaled with a 'When to Load References' section, but the body inlines methodology detail (warning system, momentum smoothing, historical confidence indicator) that the reference file already covers, so content that should be separate is inline. | 3 / 5 |
Total | 14 / 20 Passed |