CtrlK
BlogDocsLog inGet started
Tessl Logo

wagneripjr/agent-cli

Build and evaluate command-line tools for AI agent consumption: machine-readable output, semantic exit codes, hardened input, and a 0-21 agent-friendliness rubric

92

1.84x
Quality

91%

Does it follow best practices?

Impact

96%

1.84x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No findings from the security scan

Overview
Quality
Evals
Security
Files

Quality

Content

78%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is a well-structured, dense overview that routes detail into eight real one-level references with strong navigation and concrete schemas/exit codes. Its main weakness is mild redundancy between the two reference-navigation tables and validation that is gated rather than spelled out as explicit per-phase feedback loops.

Suggestions

Collapse the Quick Decision Guide and Reference Files tables into a single navigation table to remove the overlap on the eight references, or differentiate them sharply so each row earns its place.

Add an explicit validate→fix→retry checkpoint inside the build phases (e.g. after Phase 2 run the Phase 7 rubric on output architecture, then fix lowest axes before proceeding) to lift workflow clarity toward the top anchor.

Inline one minimal executable snippet per phase (e.g. a 3-line JSON envelope example in Phase 2, a TTY detection one-liner in Phase 2) so the body is copy-paste ready without requiring the framework-patterns reference.

DimensionReasoningScore

Conciseness

The body is information-dense and assumes Claude's competence (no explanations of what a CLI or JSON is), but the Quick Decision Guide table and the Reference Files table overlap noticeably in their coverage of the same eight references, a minor redundancy that could be trimmed.

4 / 5

Actionability

Guidance is concrete and specific — exact envelope fields {status, data, error, meta}, semantic exit codes (2/3/4/75), signal exits (130/143/141), explicit flag names — and defers executable code to framework-patterns.md, which is appropriate for an instruction skill; only minor gaps remain before copy-paste-ready completeness.

4 / 5

Workflow Clarity

Phases 1–6 (build) and Phase 7 (evaluate) form a clear sequence, and the Hard Gates table provides halting checkpoints plus a Common Mistakes recovery table; validation is largely enforced via gates and the post-build scoring step rather than explicit per-phase validate→fix→retry feedback loops, keeping it just below the top anchor.

4 / 5

Progressive Disclosure

SKILL.md is a clear overview pointing to eight one-level-deep reference files, every referenced path exists on disk, and navigation is doubly signaled (Quick Decision Guide by need, Reference Files by file) making discovery easy with no nested-reference chains.

5 / 5

Total

17

/

20

Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is exemplary: it states concrete capabilities, gives an explicit 'Use when'/'Triggers on' activation clause with many natural keywords, and adds sharp negative boundary guidance to avoid misfires. It uses third-person/infinitive voice throughout with no fluff.

DimensionReasoningScore

Specificity

Lists many concrete actions — 'structures commands with machine-readable JSON output on stdout', 'adds --json/--fields/--dry-run/--quiet flags', 'implements semantic exit codes', 'hardens input against path traversal and injection', 'scores CLI agent-friendliness on a 0-21 rubric across 7 axes' — giving comprehensive coverage rather than vague abstractions.

5 / 5

Completeness

It explicitly answers both what (the enumerated concrete capabilities) and when ('Use when building, designing, reviewing, or evaluating command-line tools...' plus the 'Triggers on:' list), with concrete trigger phrases as required by the top anchor.

5 / 5

Trigger Term Quality

An explicit 'Triggers on:' clause supplies numerous natural phrases users would say ('build a CLI', 'design CLI commands', 'machine-readable output', 'agent DX', 'evaluate CLI', 'score this CLI') plus synonyms and flag names, matching the comprehensive-coverage anchor.

5 / 5

Distinctiveness Conflict Risk

A clear niche (agent-friendly CLI design/evaluation) with explicit negative boundary guidance — 'NOT for TUI/full-screen apps', 'NOT for GUI apps', 'NOT for REST/GraphQL API design' — minimizes overlap with neighboring skills.

5 / 5

Total

20

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Reviewed

Table of Contents