guidance

Control LLM output with regex and grammars, guarantee valid JSON/XML/code generation, enforce structured formats, and build multi-step workflows with Guidance - Microsoft Research's constrained generation framework

Quality

73%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

Fix and improve this skill with Tessl

tessl review fix ./optional-skills/mlops/guidance/SKILL.md

Quality

Content

65%

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

The body is rich with executable, actionable code examples but is verbose and self-contained rather than acting as an overview that delegates detail to its (existing) reference files. Multi-step patterns lack validation checkpoints.

Suggestions

Trim marketing/padded sections (GitHub stars, performance percentages, the alternatives comparison table) or move them to a reference, and remove the duplicated ReAct/generate_person examples.

Replace the non-functional pseudo-grammar in the Grammar-Based Generation example with real executable guidance grammar syntax.

Move the five full 'Common Patterns' walkthroughs into references/examples.md, keeping SKILL.md as a concise overview that links to constraints.md/backends.md/examples.md inline where each topic is introduced.

Add explicit validation/checkpoint steps to the ReAct and chain-of-thought workflows (e.g., verify tool output before proceeding, guard against infinite loops).

Dimension	Reasoning	Score
Conciseness	Mostly code-driven and useful, but padded with marketing/unnecessary context ('18,000+ stars', '30-50% faster'), a comparison table, performance characteristics, and duplicated examples (ReAct agent and generate_person each appear twice), so it could be tightened.	2 / 3
Actionability	Abundant concrete, copy-paste-ready code across installation, regex/select constraints, context managers, backend configs, and common patterns — meeting the fully-executable-examples anchor.	3 / 3
Workflow Clarity	Multi-step examples (ReAct, chain-of-thought) show a sequence but lack explicit validation checkpoints or error-recovery feedback loops, leaving workflow guidance at the steps-listed-but-checkpoints-missing level.	2 / 3
Progressive Disclosure	Real one-level-deep references exist (constraints.md, backends.md, examples.md, confirmed present) and are listed in 'See Also', but the SKILL.md body itself is a ~560-line monolith containing five full pattern walkthroughs, a comparison table, and performance notes that could live in those references.	2 / 3
	Total	9 / 12 Passed

Description

82%

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 specific and distinctive, naming concrete capabilities and a clear framework niche. Its main weakness is the absence of an explicit 'Use when...' trigger clause, which caps completeness at 2.

Suggestions

Add an explicit 'Use when...' trigger clause, e.g. 'Use when you need guaranteed valid JSON/XML/code, regex- or grammar-constrained LLM output, or token-level format enforcement'.

Include a few common user phrasings ('force valid JSON', 'constrain model output', 'guaranteed structured output') to broaden natural trigger coverage.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'Control LLM output with regex and grammars', 'guarantee valid JSON/XML/code generation', 'enforce structured formats', 'build multi-step workflows' — matching the multiple-specific-actions anchor.	3 / 3
Completeness	Clearly answers what the skill does, but there is no 'Use when...' clause or equivalent explicit trigger guidance, so completeness is capped at 2 per the rubric guideline.	2 / 3
Trigger Term Quality	Uses natural terms a user would say — regex, grammars, JSON, XML, code generation, structured formats, multi-step workflows — giving good coverage of how practitioners actually phrase these needs.	3 / 3
Distinctiveness Conflict Risk	Names a specific framework (Guidance / Microsoft Research's constrained generation framework) with a clear niche in regex/grammar-constrained LLM generation, making it unlikely to trigger for the wrong skill.	3 / 3
	Total	11 / 12 Passed

Validation

75%

Warnings & errors only

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

Validation — 12 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (577 lines); consider splitting into references/ and linking	Warning
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	12 / 16 Passed

Repository: NousResearch/hermes-agent
Path: optional-skills/mlops/guidance/SKILL.md
Commit: 4c9628e

Reviewed: about 14 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.