Content
75%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 highly actionable, code-first reference that assumes Claude's competence and gives real, executable nx CLI and jq commands organized by task. The main quality issues are redundancy — flags and jq examples repeated across three sections plus a duplicated (and mislabeled) schema-pointer sentence — and a couple of shallow 'common patterns' that don't fully answer the questions they head.
Suggestions
Deduplicate: keep the full flag catalog in 'Listing Projects' only, and have 'Affected Projects' and 'Common Exploration Patterns' reference it instead of repeating the same commands.
Fix the second schema-pointer sentence (line 98): it says 'project schema' but links node_modules/nx/schemas/nx-schema.json — correct it to 'workspace/nx schema' and keep just one copy.
Strengthen the 'Why is project X affected?' pattern with the actual mechanism (e.g., nx graph --file or checking project dependencies) rather than only git diff plus the project root.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The command-per-line style is efficient and assumes Claude's competence (no Nx tutorials or concept explanations), but there is noticeable duplication: the affected/filter flags appear in both 'Listing Projects' and 'Affected Projects', the jq target examples appear in both 'Project Configuration' and 'Target Information', the 'Common Exploration Patterns' repeat commands a third time, and the schema-pointer sentence appears twice (line 52 and line 98) — the second copy is even mislabeled, saying 'project schema' while pointing to nx-schema.json. This fits 'mostly efficient but includes some unnecessary explanation or could be tightened' better than the level-4 'minor instances' anchor. | 3 / 5 |
Actionability | Every section gives fully executable, copy-paste-ready commands with real flags (--type, --withTarget, --affected --base=main, --uncommitted, jq filters over the --json output), and the 'Common Exploration Patterns' section maps common questions to concrete command pairs, including a practical grep fallback for the 'dependents' question Nx has no command for. Minor placeholders (my-app, @myorg/Y) are clearly illustrative, matching the top anchor's coverage of common cases. | 5 / 5 |
Workflow Clarity | This is a read-only exploration skill, so no destructive/batch validation cap applies, and the 'Common Exploration Patterns' section provides clear question-to-commands sequences. It stops short of a 5 because some patterns are thin or indirect — e.g., 'Why is project X affected?' only suggests git diff plus checking X's root, which doesn't actually explain why X is affected (no nx graph/dep-graph step), and the discovery path (lockfile check for package manager) is mentioned only in passing without being woven into any workflow. | 4 / 5 |
Progressive Disclosure | The body is well-sectioned with clear headers (Listing Projects, Project Configuration, Targets, Workspace Configuration, Affected, Patterns) and no external bundle files exist, so there are no nested or buried references to penalize. It scores 4 rather than 5 because at ~180 lines some reference material — the exhaustive flag catalog and the key nx.json section notes — is inline where a references/ file (or the already-pointed-to nx schemas) would keep SKILL.md a leaner overview, and the repeated schema-pointer sentences blur the navigation slightly. | 4 / 5 |
Total | 16 / 20 Passed |