anti-sycophancy

Eliminate sycophantic agreement patterns in AI responses. Load via /skill anti-sycophancy.

Quality

—

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Quality

Content

92%

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

The body is concise, concrete, and clearly sequenced with a pushback feedback loop, scoring well on instruction delivery. Its main weakness is a 'References' section that points to a README.md that does not exist in the bundle.

Suggestions

Remove or fix the dangling 'Full bibliography in README.md' reference — either add the README.md to the bundle or drop the References section since the short skill needs no external files.

Add a short before/after dialogue example to make the behavioral shift concrete and copy-paste illustrative.

Make the pushback subroutine's terminal states explicit (e.g., note when to stop restating and escalate or hold).

Dimension	Reasoning	Score
Conciseness	The body is lean (~30 lines) with no padding or explanation of concepts Claude already knows; every line is actionable instruction that earns its place. Not score 2 (no unnecessary explanation to tighten); not score 1 (not verbose).	3 / 3
Actionability	Gives concrete, specific behavioral guidance — 'State it in one sentence stripped of premises', 'Respond with the conclusion first, evidence second' — plus a complete pushback subroutine; for an instruction-only skill this is concrete and complete, not vague. Not score 2 (no missing key details); not score 1 (it instructs rather than describes).	3 / 3
Workflow Clarity	A clearly numbered 4-step sequence (Extract → Assess → Conclude → Respond) with an explicit pushback feedback loop that acts as error recovery. This is not a destructive/batch operation so the missing-validation cap does not apply. Not score 2 (sequence and feedback loop are explicit); not score 1 (steps are clearly sequenced).	3 / 3
Progressive Disclosure	Sections are well-organized for a short skill, but the 'References' section points to 'Full bibliography in README.md' and no README.md (or any bundle file) exists in the skill directory — a dangling navigation dead-end. Not score 3 (broken reference prevents clean navigation); not score 1 (good section organization, no nested references).	2 / 3
	Total	11 / 12 Passed

Description

57%

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 occupies a clear niche, but it offers only a single action and substitutes a load command for an explicit 'Use when...' trigger, leaving the 'when' essentially unstated.

Suggestions

Add an explicit 'Use when...' trigger clause (e.g., 'Use when the assistant is tempted to agree with an unsupported user claim, or when the user pushes back on a correct assessment').

Expand beyond the single 'Eliminate' verb to list the concrete behaviors the skill performs (e.g., 'challenge user claims independently, separate evidence from deference, restate position under pushback').

Soften the jargon with natural user-facing terms such as 'agreement bias' alongside 'sycophantic agreement patterns'.

Dimension	Reasoning	Score
Specificity	Names the domain ('sycophantic agreement patterns') and one concrete action ('Eliminate'), but lists only a single action rather than multiple specific capabilities. Not score 3 (no list of several concrete actions); not score 1 (it is specific, not vague).	2 / 3
Completeness	Clearly states what the skill does ('Eliminate sycophantic agreement patterns in AI responses'), but the only 'when' guidance is 'Load via /skill anti-sycophancy' — an invocation command, not a trigger condition. Per the guideline, a missing 'Use when...' clause caps completeness at 2; not score 1 because the 'what' is explicit.	2 / 3
Trigger Term Quality	Includes relevant keywords ('sycophantic', 'agreement patterns', 'AI responses') but leans toward jargon and misses common natural variations a user might say ('agreement bias', 'stop agreeing with me', 'people-pleasing'). Not score 3 (incomplete natural-term coverage); not score 1 (has some genuinely relevant keywords).	2 / 3
Distinctiveness Conflict Risk	Targets a clear, narrow behavioral niche (anti-sycophancy / agreement bias) with triggers unlikely to overlap with typical file- or code-processing skills. Not score 2 (the niche is distinctly identifiable, not broadly overlapping).	3 / 3
	Total	9 / 12 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: sickn33/antigravity-awesome-skills
Commit: 6e69f00

Reviewed: 5 days 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.