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 tightly written command-catalog skill: lean token usage, executable safe-start commands, explicit write-safety checkpoints, and clean deferral of per-command detail to runtime help and schema discovery. The main gap is the absence of any worked example of a typical operation and any post-write verification step.
Suggestions
Add one or two copy-paste-ready examples of common operations, e.g., creating a deck from markdown or inserting text by objectId.
Add a post-write verification step to the workflow, e.g., "After a write, confirm the result with `gog slides read-slide` or `info` before reporting success."
State how to handle a failed write (e.g., inspect the error, re-run with the corrected objectId/range) to close the feedback loop for destructive operations.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and dense: a compact safe-start command block, five terse safety bullets, and a command table where every row states a command and its purpose in one line. No concepts Claude already knows are explained; the only non-essential text is a one-line build-artifact comment. | 5 / 5 |
Actionability | The safe-start block ("gog auth list --check --json --no-input", "gog schema slides --json") is copy-paste ready, and the footer gives a concrete discovery contract ("Run `gog slides <command> --help`... `gog schema slides <command> --json`"). Not anchor 5 because no worked example of a common operation (e.g., an actual `create-from-markdown` or `insert-text` invocation with arguments) is shown. | 4 / 5 |
Workflow Clarity | There is a clear entry sequence (read the parent gog skill, check auth, inspect schema, explore with --readonly) and explicit write-safety checkpoints ("--dry-run before supported writes", "Confirm the exact account, object, and mutation before any write or delete"), so the destructive-operation cap does not apply. Not anchor 5 because there is no post-write verification or feedback loop (e.g., re-read the slide with `read-slide`/`info` to confirm the mutation). | 4 / 5 |
Progressive Disclosure | The single-file skill is well organized into overview sections, per-command details are delegated to runtime `--help` and machine-readable `schema` output rather than inlined, and the one file reference ("Read `../gog/SKILL.md` first for shared auth, output, safety, and live-write rules") is clearly signaled, states exactly what it covers, and is one level deep. | 5 / 5 |
Total | 18 / 20 Passed |