Content
85%Weight 40%Scale 1-3Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is highly actionable with executable commands, a clear validated workflow, and good error-recovery guidance, all well-organized in a self-contained file. The only weakness is redundant duplication of the bash command block between the build and run steps, which mildly hurts token efficiency.
Suggestions
Remove the duplicated bash command block: state the command once and have step 3 reference it (e.g. "launch the command from step 2 as a background task") to recover token efficiency.
Consider consolidating the "IMPORTANT" and "Explicit invocation only" notes, which overlap on the explicit-invocation requirement, to tighten the lead-in.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean with no concept-explaining fluff, but the full bash command block is duplicated verbatim in step 2 and step 3, which is avoidable token waste; this matches anchor 2's "mostly efficient but could be tightened" rather than anchor 3. | 2 / 3 |
Actionability | It gives fully executable bash commands (git rev-parse --verify, roborev review --branch --wait) plus a concrete Task tool invocation with run_in_background and subagent_type: "Bash", matching anchor 3's copy-paste-ready standard. | 3 / 3 |
Workflow Clarity | A clear 5-step sequence includes an explicit validation checkpoint (step 1 validates the ref and halts on failure) and error-recovery guidance in step 4, matching anchor 3's explicit validation and feedback loops. | 3 / 3 |
Progressive Disclosure | With no bundle files present the content is appropriately self-contained and organized into well-signaled sections (Usage, Instructions, Examples, See also) with no monolithic wall or nested references, matching anchor 3. | 3 / 3 |
Total | 11 / 12 Passed |