Content
92%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.
An excellent body: copy-paste-ready tool invocations, a numbered workflow with error-recovery guidance and mandatory validation steps, and a correctly split-out QC script. Only minor conciseness trims are possible (the thrice-repeated specificity caveat).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Overall lean and tool-specific — the table of design rules, command examples, and gotchas all earn their place with minimal generic explanation. Minor trimmable redundancy: genome-specificity appears three times (Step 4, gotcha #1, and Honest limitations), and the table's brief 'Why' column slightly restates knowledge Claude already has. | 4 / 5 |
Actionability | Fully executable commands with real parameters (`tu run DNA_primer_design '{...}'`, `tu run NEB_Tm_calculate` with polymerase/concentration values), a real QC script (`scripts/primer_qc.py`, verified present), exact error-message strings paired with fixes, and concrete numeric targets throughout. | 5 / 5 |
Workflow Clarity | Clearly sequenced Steps 1–5 (design → Tm/Ta → rules QC → specificity → gotchas) with explicit validation checkpoints: the constraint-quirk block maps each error message to a fix (a feedback loop), the QC script validates pairs, and BLAST confirmation is mandated before ordering. | 5 / 5 |
Progressive Disclosure | The body is a well-sectioned overview (~75 lines) with one clearly signaled, one-level-deep bundle reference (`scripts/primer_qc.py` in Step 3), which exists and does what the body claims. No buried or nested references; nothing that belongs in a separate file is inlined. | 5 / 5 |
Total | 19 / 20 Passed |