CtrlK
BlogDocsLog inGet started
Tessl Logo

pantheon-ai/challenge

Challenge AI output with structured devil's-advocate protocols: anchor, verify, framing, and deep sub-commands.

65

Quality

82%

Does it follow best practices?

Impact

No eval scenarios have been run

SecuritybySnyk

Passed

No known issues

Overview
Quality
Evals
Security
Files

Quality

Content

62%

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

A well-organized instruction skill with a clear dispatch flow, explicit validation guard, and structured output template. Its main weaknesses are redundant When-to-Use text, and that the protocol/reference files it dispatches to are missing from the bundle, undermining both actionability and progressive disclosure.

Suggestions

Trim or remove the 'When to Use' section — its triggers duplicate the frontmatter description's 'Use when' clause, adding tokens without new information.

Bundle the referenced protocol files (references/protocols/anchor.md, framing.md, verify.md and references/reference.md) or inline their core execution steps, since the body dispatches to files that are absent and cannot stand alone.

Verify all cited bundle paths resolve: references/reference.md and references/protocols/*.md are referenced in the Dispatch table and References section but do not exist in the bundle, breaking navigation.

DimensionReasoningScore

Conciseness

Mostly efficient and free of concepts Claude already knows, but the "When to Use" section restates triggers already in the frontmatter description, and the five-entry Anti-Patterns block with per-item "Why:" rationales adds tokens that could be tightened.

2 / 3

Actionability

Dispatch table, Agent-spawn spec, and 4-part Thinking Transparency structure are concrete, but the actual protocol execution is deferred to references/protocols/*.md files that are absent from the bundle, leaving the executable core incomplete.

2 / 3

Workflow Clarity

Clear dispatch sequence (parse first word → subcommand → read protocol → execute), an explicit validation checkpoint in the AskUserQuestion empty-answers guard with a fallback feedback loop, and a checklist-style output structure.

3 / 3

Progressive Disclosure

The overview is well-structured with clearly signaled one-level-deep references and a descriptive References section, but the cited files (references/reference.md, references/protocols/{anchor,framing,verify}.md) are not present in the bundle, so the disclosure is non-functional.

2 / 3

Total

9

/

12

Passed

Description

100%

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

An excellent description: third-person, concise, with an explicit Use-when clause packed with natural trigger terms and four named subcommands mapping to concrete error types. It clearly answers both what and when with low conflict risk.

DimensionReasoningScore

Specificity

Names concrete actions ("Challenge, push back, play devil's advocate on AI output") and enumerates four specific subcommands each tied to a distinct error type (anchor/verify/framing/deep), matching the multiple-specific-actions anchor.

3 / 3

Completeness

It explicitly states what the skill does (challenge/push back/devil's advocate on AI output, plus subcommands) and when to use it via an explicit "Use when:" clause with concrete triggers, satisfying both what and when.

3 / 3

Trigger Term Quality

The "Use when:" clause lists ~12 natural phrases a user would actually say ("are you sure", "prove it", "what if you're wrong", "poke holes", "second opinion", "sanity check", "really?"), giving strong coverage of natural trigger terms.

3 / 3

Distinctiveness Conflict Risk

The challenge/devil's-advocate niche with triggers like "prove it", "poke holes", and "what if you're wrong" is clearly distinct from other skills; voice is third person with no first/second-person phrasing.

3 / 3

Total

12

/

12

Passed

Validation

75%

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

Validation12 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

allowed_tools_field

'allowed-tools' contains unusual tool name(s)

Warning

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

relative_links

Relative link issues: 4 missing, 3 deeper-than-1-level

Warning

referenced_paths_exist

Referenced path issues: 9 missing, 6 deeper-than-1-level

Warning

Total

12

/

16

Passed

Reviewed

Table of Contents