Content
53%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 is well-organized with tagged sections, a real asset reference for IaC details, and several concrete constraints, but it reads as a principles list rather than an executable workflow — no sequenced implementation steps, no feedback loops, and vague mandates mixed in with the concrete ones. It sits near the middle of the scale across all dimensions.
Suggestions
Restructure the core workflow as an ordered sequence (search existing code → implement → validate in dependency order → review git changes) with explicit validate-then-fix-then-revalidate checkpoints.
Make key directives executable: specify how to verify multi-environment configuration and what "check logs and running services" concretely means, or move detail to a reference file.
Trim filler mandates ("Zero tolerance: no cheating, no pre-existing excuses") and tighten the run-on background-terminal bullet into concrete script steps or a scripts/ file.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly terse bullet lists that assume competence, but it includes run-on mandates (the background-terminal script bullet), vague filler ("Zero tolerance: no cheating, no pre-existing excuses"), and a when-to-use section that merely repeats the description. | 3 / 5 |
Actionability | Some genuinely concrete guidance exists (validation order databases → APIs → Web → Mobile, the CONTEXT.md file list, the 300 LOC limit, 5-15 second timeouts, named MCPs), but most directives are high-level with no commands or code and unclear referents like "Check logs and running services locally". | 3 / 5 |
Workflow Clarity | A dependency-ordered validation sequence and a validation checklist are present, but there is no ordered implementation workflow with embedded checkpoints, and validation lacks a fix → re-validate feedback loop. | 3 / 5 |
Progressive Disclosure | The body is a reasonable overview with a clearly signaled, real one-level-deep asset reference (`assets/iac.md`, 235 lines), though it uses XML-style tags instead of markdown headers and the IaC material could be more discoverable with a dedicated resources section. | 4 / 5 |
Total | 13 / 20 Passed |