mrkeyoor.com_
Thu 13 Aug 15:38 UTC
LLM Toolsevaluationupdated 13 Aug 2026

skill-up

Skill Up is a command-line test runner for Agent Skills used by Codex, Claude Code, Qoder CLI, and Qwen Code. It runs repeatable cases inside those real agent engines, grades the results, and pairs with an included skill that can help diagnose failures and improve both the target skill and its test suite.

Verdict

Skill Up is the most focused option here for teams treating Agent Skills as software that deserves repeatable tests. The real-client execution, baseline comparison, varied judges, and CI reports are more useful than another generic prompt scorer. Adopt it for an actively maintained skill suite, but pin the tool, inspect raw artifacts when a judge fails strangely, and expect to work around a few cross-engine and workspace limitations while the project matures.

Setup3/5Quick helper install, but credible suites need agent and judge setup
Docs4/5Clear bilingual manual, schemas, examples, and CI guidance
Community3/5Small but highly active project with responsive pull-request work
Maturity3/5Useful releases, with important cross-engine gaps still open

Who it’s for

Agent Skill authors who need regression tests beyond manually prompting an assistant.
Teams supporting the same skill in Codex, Claude Code, Qoder CLI, or Qwen Code.
Maintainers who want YAML test cases, JUnit or HTML output, and pull-request checks.
Developers migrating Anthropic-style evals.json suites into a richer local workflow.

Who it’s NOT for

Teams whose main problem is skill discovery or routing: issue #166 says positive and negative routing evaluation is still a proposed capability.
Codex users whose judge sessions can emit very large command-output records: issue #172 reports silent truncation above 1 MiB, although a fix is under review.
Evaluators mixing a Codex case engine with an Anthropic judge model: issue #104 says the judge inherits the case engine and the protocol mismatch can fail.
Native Windows teams expecting full parity: the README points to Windows-specific limitations, and open work covers native agent CLI tests and Windows OpenSandbox guests.
Local evaluations that must start inside an existing repository without prompt-based directory changes: issue #119 says the none runtime creates a temporary workspace and lacks an explicit host-workspace option.

Setup reality

Installing the conversational skill-upper helper is one npx skills add command, and it can guide the agent through installing the Go CLI. Manual users can run the shell installer or build with Go 1.25 or later, then create an eval.yaml plus case files. The first useful suite is still real test engineering: choose with-skill and baseline cases, configure agent credentials, control fixtures and workspaces, select rule, script, or model judges, and decide what a meaningful pass means. CI adds a Linux Docker action, secret management, model cost, and careful version pinning because the action image's bundled CLI does not automatically match every CLI release tag.

Agent Skills need tests, not good intentions

A SKILL.md file can look perfectly sensible and still fail when a real agent interprets it. Instructions may trigger at the wrong moment, conflict with tool behavior, miss a workspace assumption, or produce an answer that appears plausible while skipping a required step. Skill Up treats that problem like software testing. Developers describe cases in YAML, run them through supported agent clients, grade the outcomes, and retain structured results for comparison and CI.

The project supports Codex, Claude Code, Qoder CLI, and Qwen Code as built-in engines, plus custom local agents. A case can be judged by straightforward rules, a script, or another agent. Reports include JSON, Markdown, HTML, JUnit XML, and Anthropic-compatible grading and benchmark files. That range matters because a local author wants readable failure evidence, while a pull-request job needs a machine-readable exit condition.

Skill Up also runs with and without the target skill, following the evaluation pattern documented by the Agent Skills specification. That baseline is essential. A passing answer proves little if the underlying agent succeeds just as often without the skill. The useful question is whether installing the skill improves the behavior the author intended.

The evolution loop is the distinctive part

The repository includes skill-upper, an Agent Skill that drives the evaluator through conversation. Ask it to inspect a target SKILL.md, identify important behavior, generate cases, run them, interpret failures, and improve either the skill or its tests. Subsequent failures can become regression cases rather than disappearing into a chat transcript.

This is a smart interface for a configuration-heavy tool. Authors can start from intent instead of memorizing every YAML field. The helper also checks whether the underlying CLI exists and guides installation. Codex and Claude Code users can add it globally with the skills package command, then work inside the target project.

Automation introduces a judgment risk, however. A passing suite is not automatically a good suite, especially when the same model family helps write the instruction, generate the test, and grade the result. Developers still need to review whether cases represent real requests, whether failures indicate a broken skill or a broken expectation, and whether an agent judge received enough independent evidence. Rule and script judges are preferable when correctness can be stated deterministically. Model judges belong where semantics genuinely resist exact checks.

Manual use is approachable but not trivial

The binary installer downloads a release for the current platform, while source builders need Go 1.25 or newer. A minimal suite consists of an eval.yaml file and one case file. Defaults keep the first run short: one worker, ten turns, a five-minute timeout, and JSON output. When the eval folder sits beside SKILL.md, Skill Up automatically installs the local skill for the run. A separate validate command catches configuration mistakes before model calls begin.

Real suites quickly add more decisions. Authors must provision the chosen agent CLI and credentials, create fixtures, manage temporary workspaces, separate output artifacts, and control parallelism against API limits and cost. The none environment is useful for plain-text skills, but issue #119 explains that it starts in a generated temporary directory rather than an explicitly selected existing repository. Long end-to-end cases may have to repeatedly tell the model where to change directories, a fragile workaround when repository context is the subject of the test.

CI is credible rather than decorative. The included GitHub Action runs in a Linux container with agent CLIs prebuilt, accepts model and provider settings, and can test more than one engine. Its versioning deserves close reading. The runner image pins a Skill Up binary and uses an immutable image digest, but a new CLI release does not automatically refresh that image. Older release tags may predate the action entirely. Teams wanting reproducibility should pin a known post-refresh commit or verified tag and inspect the version printed in logs.

Cross-engine support has sharp edges

Supporting several autonomous clients is Skill Up's biggest advantage and its largest maintenance burden. Release v0.9.0 fixed session lookup, cross-workspace resume protection, multi-turn response scoping, and the required CI gate. Those are encouraging fixes, but they show how much client-specific transcript and session behavior the tool must normalize.

Two open bugs are especially relevant. Issue #172 reports that Codex JSONL parsing stops silently when one record exceeds 1 MiB, potentially leaving an earlier progress message as the supposed final result. A large search output inside an agent-judge run can therefore produce a misleading grading error even when valid final JSON exists later. Pull request #184 is working on that defect, but users should still inspect raw output when Codex judging fails unexpectedly.

Issue #104 covers a different boundary: the agent judge inherits the case engine even when its configured model uses another provider's protocol. A Codex evaluation paired with an Anthropic judge can fail before meaningful grading. Use a protocol-compatible judge or test the exact engine and model pairing before standardizing a suite.

Skill routing is another unfinished area. The current evaluator measures task performance once a skill is installed. Issue #166 proposes tests for whether a skill should have triggered, should have stayed quiet, or shadowed another skill. Until that exists, teams need a separate way to test descriptions and activation behavior.

Young, active, and already useful

Skill Up was pushed on August 13, 2026, one day after v0.9.0, with active pull requests addressing parser and documentation work. The open count of 16 combines issues and PRs, so it represents a small live queue rather than 16 confirmed bugs. Releases have arrived rapidly since May, which signals energy but also a moving interface. Pinning matters.

The English README is the default, and a Chinese README and bilingual user manual are available. Documentation covers the quick start, configuration, cases, judges, reports, custom engines, CI, Windows, and Anthropic migration. Apache 2.0 licensing is friendly for company use.

For dedicated Agent Skill development, Skill Up is easier to recommend than a generic evaluation framework because it owns the awkward parts: installing the skill, invoking real coding agents, capturing their work, and comparing a baseline. Promptfoo is broader, OpenAI Evals is better aligned with model benchmarks, and OpenEvals supplies evaluator components without this execution loop. Choose Skill Up when the thing under test is specifically an Agent Skill, then keep humans responsible for the quality of the tests and the meaning of a pass.

Alternatives

ProjectWhat it isPick it when
PromptfooA broad prompt, agent, RAG, and red-team evaluation framework with declarative tests.pick this instead when you need provider-wide prompt comparisons or security testing more than native Agent Skill installation and execution.
OpenAI EvalsA framework and benchmark registry for evaluating language models and model systems.pick this instead when you are building model-level benchmarks rather than testing a SKILL.md inside interactive coding agents.
OpenEvalsA library of reusable evaluators for scoring outputs from language-model applications.pick this instead when your application already owns the execution harness and mainly needs evaluator building blocks.

What people are saying

  1. [github-trending] alibaba/skill-up

Sources

  1. Skill Up README
  2. Skill Up v0.9.0 release notes
  3. Codex JSONL truncation issue
  4. Cross-provider agent judge issue
  5. Host workspace support issue
  6. Routing evaluation proposal