Content
86%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: executable commands throughout, a numbered core workflow with precheck and observation steps, and clean single-file organization. The main improvements are deduplicating the repeated session-creation lines and adding error-recovery guidance.
Suggestions
Deduplicate the repeated session-creation pipeline: state the AGENT_HOME/SESSION_ID bootstrap once and reference it from the Common Patterns recipes to cut redundant tokens (conciseness).
Add brief error-recovery guidance after the doctor and wait steps — e.g., what to inspect or do when 'doctor --json' reports missing prerequisites or a 'wait' times out — to close the feedback-loop gap (workflow_clarity).
Consider consolidating the Core Workflow example with the overlapping Essential Commands block, since both enumerate the same command set.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean, assumes Claude's competence, and explains no known concepts, but the session-creation pipeline 'agent-terminal --home ... create --json -- /bin/bash | jq -r '.result.sessionId'' is repeated verbatim five times across sections — minor redundancy that could be trimmed, so anchor 4 rather than 5. | 4 / 5 |
Actionability | Every example is fully executable copy-paste shell with consistent --json usage, covering lifecycle, interactive TUI driving, and artifact export — matching the 'fully executable, copy-paste ready, covers common cases' anchor; the '<interactive-command>' placeholder is justified flexibility, not pseudocode. | 5 / 5 |
Workflow Clarity | The 8-step Core Workflow is clearly sequenced with an explicit 'doctor --json' precheck and wait-based observation checkpoints, but there is no error-recovery feedback loop (e.g., what to do when doctor fails or wait times out), fitting anchor 4 ('most checkpoints present; minor validation gaps') rather than 5. | 4 / 5 |
Progressive Disclosure | No bundle files exist and the skill is a single well-sectioned file with no nested or buried references; all content earns its place and navigation is trivial, so it matches the clear, easy-to-navigate top anchor for a self-contained skill. | 5 / 5 |
Total | 18 / 20 Passed |